CHAPTER-1

INTRODUCTION

Flood Determining can be characterized as a course of assessment of time and term

in light of geographical qualities of any stream bowl which decreased the perils to human
existence and climate too. Flood Estimating method challenge to foresee event and greatness
with season of glimmer flood. Flood occurred because of proceed with precipitation with
particular time. Conventional precipitation additionally adds to change with time into lethal
flood. Flood estimating procedures assume a crucial and significant part to relieve the perils
for non-primary designs with financially savvy the executives. Flood anticipating stations
cover the organization of flood inclined region to give flood advance notice to organization.
Determining inflow additionally utilized for the activity of pressure driven designs, for
example, dam on which opening and shutting of entryways on spillways. Flood guaging
methods and Flood cautioning framework are required different sort of designs of flood. Flood
might be diminished by developing designs like dams, weirs, dykes yet can't take out the
gamble. Flood determining methods ready to moderate the risks for populace and climate
progressively with an early advance notice [1]. Flood anticipating has been approach through
precipitation - overflow and flood directing model. Flood conjecture foresee inflow at chose
area with HFL esteem at chosen areas of stream with time relies upon watershed or catchment
region. Afterward, downstream side predicts the flood restricted with movement time with an
evaluation of vulnerabilities to appropriately uphold the chiefs exercises [9]. Flood
anticipating utilizing AI calculation (MLA) technique comprehend to learn and further
develop framework scale to relieve flood perils as per the environmental change with the
assistance of simulated intelligence. Flood determining utilized for making the AI calculation
with past and genuine records of flood with ongoing information utilizing precipitation
measures for various returning time. The dataset sources are precipitation spillover, water
levels utilizing programmed downpour checks with satellites innovation, invasion rate and so
on.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 1
 CHAPTER-2
LITERATURE SURVEY
Successful catastrophe risk the board depends on science-based answers for close the
hole among anticipation and readiness measures. The meeting on the Unified Countries
post2015 structure for catastrophe risk decrease features the requirement for cross-line early
admonition frameworks to reinforce the readiness periods of calamity risk the board, to save
lives and property and diminish the general effect of serious occasions. Mainland and
worldwide scale flood estimating frameworks give fundamental early flood cautioning data to
public and global common security specialists, who can utilize this data to go with choices on
the most proficient method to plan for impending floods. Here the expected financial
advantages of early flood alerts are assessed in view of the conjectures of the mainland scale
European Flood Mindfulness Framework (EFAS) utilizing existing flood harm cost data and
estimations of potential stayed away from flood harms. The advantages are of the request for
400 Euro for each 1 Euro contributed. A responsiveness examination is acted to test the
vulnerability in the technique and foster an envelope of possible money related advantages of
EFAS alerts. The outcomes give obvious proof that there is logical a significant financial
advantage in this cross-line mainland scale flood early admonition framework. This supports
the more extensive drive to carry out early advance notice frameworks at the mainland or
worldwide scale to work on our versatility to normal perils.

Measurable examination of the effect of radar precipitation vulnerabilities on water assets

Creator: He, X.; Refsgaard, J.C.; Sonnenborg, T.O.; Vejen, F.; Jensen, K.H Conceptual:

Vulnerability examination in hydrological demonstrating has turned into a fundamental

stage in the logical translation of model outcomes and a helpful device to help navigation.
Among numerous vulnerability sources in the demonstrating practice, vulnerabilities in
precipitation assessment assume a significant part since it is the super main thrust for other
hydrological processes. The current review exhibits a factual strategy for creating radar
precipitation acknowledge that record for the vulnerabilities in radar-based quantitative
precipitation assessment (QPE)

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 2
 CHAPTER-3
SYSTEM ANALYSIS

3.1 EXISTING SYSTEM:

The current review completed on Wardha Waterway bowl .The Upper Wardha project
is one of the significant water system project in Vidharbha locale of Maharashtra state. This
task is across Wardha Stream close to town Simbhora Taluka Morshi of Amravati area. The
Upper Wardha project comprises of earthen dam on a the two flanks with a midway found
gated spillway and waterway on left and right flanks. The Upper Wardha dam has absolute
grass water system (GI) of 11690 (ha) with gross limit of 786 MCM in Godavari Stream Bowl
with yearly precipitation in the catchment is 840 mm\

DISADVANTAGES OF EXISTING SYSTEM:

1) Less precision
2) low Effectiveness

3.2 PROPOSED SYSTEM

Flood determining procedure coordinated technique reasonably founded on accessible

information and an evaluation of rating measures with a motivated presentation. Flood
guaging involving continuous assessment gives chances of flood esteem in GUI. Flood
assessment utilizing AI continuously can compute huge information quickly. Correlation
between flood demonstrating by AI and stochastic technique (for example Muskinghum
strategy) gives AI is precise, simple and can be applied for quantities of estimation.

3.2.1 Advantages of proposed system

1) High precision
2) High productivity

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 3
 3.3 FEASIBILITY STUDY

SYSTEM STUDY FEASIBILITY STUDY

The practicality of the task is examined in this stage and strategic agreement is advanced
with an extremely broad arrangement for the undertaking and a few quotes. During framework
examination the plausibility investigation of the proposed framework is to be done.

Three key contemplations engaged with the achievability investigation are

• Conservative Possibility

• Specialized Plausibility

• Social Plausibility

Conservative Achievability
This study is completed to check the monetary effect that the framework will have on
the association. How much asset that the organization can fill the innovative work of the
framework is restricted. The uses should be legitimate. Accordingly the created framework
also affordable and this was accomplished in light of the fact that the greater part of the
advances utilized are openly accessible. Just the modified items must be bought.

Specialized Feasibility
This study is completed to check the specialized possibility, or at least, the
specialized necessities of the framework. Any framework created should not have a popularity
on the accessible specialized assets. This will prompt high requests on the accessible
specialized assets. This will prompt high requests being put on the client. The created
framework should have an unassuming prerequisite, as just insignificant or invalid changes
are expected for executing this framework.

Social Attainability

The part of study is to actually take a look at the degree of acknowledgment of the
framework by the client. This incorporates the method involved with preparing the client to
productively utilize the framework.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 4
 CHAPTER-4
SYSTEM REQUIREMENTS

4.1 EQUIPMENT AND PROGRAMMIMG PREREQUISITES:

Hardware Requirments :
• framework : i3 or above.

• slam : 4 gb.

• hard circle : 40 gb

Programming Requirments :
• working framework : windows8 or above.

• coding language : python

• front end : html/css

• backend : my sql

• server : django

Functional Requirements: information assortment: assemble applicable information

sources including climate information, waterway levels, territory data, and authentic flood
information. preprocessing: clean and preprocess the gathered information to guarantee
consistency, exactness, and similarity for ai calculations. include choice: distinguish and
choose pertinent elements from the dataset that are characteristic of flood events and can add
to exact estimating. model turn of events: foster ai models, like relapse, grouping, or time
series models, to anticipate flood events in light of the chose highlights. preparing and
approval: train the ai models utilizing authentic information and approve their presentation
utilizing suitable assessment measurements and approval methods. constant expectation:
carry out systems to persistently screen approaching information and give ongoing flood
estimates utilizing the prepared ai models.

4.1.1 Non-practical Prerequisites:

Non-practical prerequisite (nfr) indicates the quality trait of a product framework. they judge
the product framework in view of responsiveness, ease of use, security, movability and other
nonutilitarian principles that are basic to the progress of the product framework. illustration of
nonfunctional prerequisite, "how quick does the site load?" neglecting to meet non-practical
necessities can bring about frameworks that neglect to fulfill client needs.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 5
 • Usability requirement
• Serviceability requirement

• Manageability requirement

• Recoverability requirement

• Security requirement

• Information trustworthiness necessity

• Capacity requirement

• Availability requirement

• Scalability requirement

• Interoperability requirement

• Reliability requirement

• Maintainability requirement

• Regulatory requirement

• Environmental requirement

4.2 PURPOSE
This overview distinguishes the cutting edge of ML techniques for flood expectation
where peer-surveyed articles in high level subject fields are checked on. Among the articles
distinguished, through search questions utilizing the pursuit system, those including the
exhibition assessment and correlation of ML strategies were given need to be remembered for
the audit to recognize the ML techniques that perform better specifically applications.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 6
 CHAPTER-5
SYSTEM ARCHICTURE

5.1 SYSTEM ARCHICTURE:

Fig : 5.1 System Architecture

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 7
RELATED WORK :

The overall uses of ML in all flood asset factors according to the viewpoint of ML
displaying and information driven forecast frameworks. Regardless, ML calculations have
significant qualities that should be painstakingly thought about. The first is that they are
comparable to their preparation, by which the framework learns the objective undertaking in
view of past information. In the event that the information is scant or doesn't cover assortments
of the assignment, their learning misses the mark, and consequently, they can't perform well
when they are invested into effort. Hence, utilizing strong information improvement is
fundamental through, e.g., executing a dissemination capability of amounts of loads [56],
invariance evaluations to hold the gathering qualities [57], or recuperating the missing factors
utilizing causally subordinate coefficients [58]. The subsequent perspective is the capacity of
every ML calculation, which might shift across various kinds of undertakings. This can
likewise be known as a "speculation issue", which shows how well the prepared framework
can foresee cases it was not prepared for, i.e., whether it can foresee past the scope of the
preparation dataset.

These attributes of the calculations should be explained as for the sort and measure
of accessible preparation information, and the kind of expectation task, e.g., water level and
streamflow. In this audit, we investigate instances of the utilization of different ML
calculations for different sorts of undertakings. At the theoretical level, we chose to partition
the objective errands into present moment and long haul forecast. We then, at that point,
investigated ML applications for flood-related undertakings, where we organized ML
strategies as single techniques and cross breed strategies. Cross breed strategies are those that
join more than one ML strategy. Here, we ought to take note of that this paper overviews ML
models utilized for expectations of floods on destinations where downpour checks or
insightful detecting frameworks utilized. Our objective was to review forecast models with
different lead times to floods at a specific site. According to this point of view, spatial flood
expectation was not engaged with this review, as we didn't concentrate on forecast models
used to appraise/distinguish the area of floods.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 8
 CHAPTER-6
SYSTEM DESIGN

6.1 UML DIAGRAMS :

UML represents Brought together Demonstrating Language. UML is a normalized
universally useful demonstrating language in the field of item situated programming. The
standard is made due, and was made by, the Article The board Gathering.
The objective is for UML to turn into a typical language for making models of item situated
PC programming. In its ongoing structure UML is contained two significant parts: a
Metamodel and a documentation. Later on, some type of strategy or process may likewise be
added to; or related with, UML.

The Brought together Demonstrating Language is a standard language for determining,

Perception, Developing and recording the relics of programming framework, as well
concerning business displaying and other non-programming frameworks.
The UML addresses an assortment of best designing practices that have demonstrated
fruitful in the displaying of huge and complex frameworks. The UML is a vital piece of
creating objects situated programming and the product improvement process. The UML
utilizes generally graphical documentations to communicate the plan of programming
projects.

Objectives:
The Essential objectives in the plan of the UML are as per the following:

1. Provide clients a prepared to-utilize, expressive visual demonstrating Language with the
goal that they can create and trade significant models.
2. Provide extendibility and specialization systems to expand the center ideas.
3. Be free of specific programming dialects and advancement process.
4. Provide a proper reason for figuring out the displaying language.
5. Encourage the development of OO instruments market.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 9
 6.1.1 Use case diagram:

A utilization case graph in the Bound together Displaying Language (UML) is a sort of
conduct chart characterized by and made from a Utilization case examination. Its motivation
is to introduce a graphical outline of the usefulness given by a framework concerning
entertainers, their objectives (addressed as use cases), and any conditions between those
utilization cases. The principal motivation behind a utilization case chart is to show what
framework capabilities are performed for which entertainer. Jobs of the entertainers in the
framework can be portrayed.

proprocess dataset
home

user login

run machinelearing alogrithm

User login
User sigup

new user sigup here frocast flood

logot

Fig : 6.1.1 Use case diagram

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 10
 6.1.2 Class diagram:

In computer programming, a class graph in the Brought together Demonstrating Language

(UML) is a kind of static design chart that portrays the construction of a framework by
showing the framework's classes, their properties, tasks (or strategies), and the connections
among the classes. It makes sense of which class contains data

Fig : 6.1.2 Class Diagram

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 11
6.1.3 Sequence diagram:
A succession graph in Brought together Demonstrating Language (UML) is a sort of
collaboration chart that shows how cycles work with each other and in what request. It is a
build of a Message Succession Graph. Grouping graphs are now and then called occasion
charts, occasion situations, and timing outlines.

Fig : 6.1.3 Sequence Diagram

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 12
 6.2 COLLABARATION DIAGRAM:
Action charts are graphical portrayals of work processes of stepwise exercises and activities
with help for decision, emphasis and simultaneousness. In the Bound together Displaying
Language, action graphs can be utilized to depict thebusiness and functional bit by bit work
processes of parts in a framework. An action outline shows the general progression of control.

Fig : 6.2 Collabaration Diagram

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 13
 CHAPTER-7
SOFTWARE ENVIRONMENT

7.1. SOFTWARE ENVIRONMENT

What is Python: The following are a few realities about Python.

Python is presently the most broadly utilized multi-reason, significant level programming

language.

Python permits programming in Item Arranged and Procedural ideal models. Python

programs for the most part are more modest than other programming dialects like Java.

Developers need to type generally less and space prerequisite of the language, makes them

discernible constantly.

The greatest strength of Python is immense assortment of standard library which can be

utilized for the accompanying -

• AI

• GUI Applications (like Kivy, Tkinter, PyQt and so forth )

• Web systems like Django (utilized by YouTube, Instagram, Dropbox)

• Picture handling (like Opencv, Cushion)

• Web scratching (like Scrapy, BeautifulSoup, Selenium)

• Test structures

• Sight and sound

7.1.1 Advantages of Python :-

1. Broad Libraries
2. Embeddable
3. Further developed Efficiency
4. IOT Open doors
5. Coherent
6. Object-Situated
7. Free and Open-Source
8. Convenient
9. Deciphered

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 14
Advantages of Python Over Other Languages :

1. Less Coding
2. Reasonable
3. Python is for Everybody
Impediments of Python
Up to this point, we've seen the reason why Python is an extraordinary decision for your
task. Yet, in the event that you pick it, you ought to know about its ramifications also. How
about we currently see the disadvantages of picking Python over another dialect.

1. Speed Constraints

We have seen that Python code is executed line by line. However, since Python is
deciphered, it frequently brings about sluggish execution. This, in any case, isn't an issue
except if speed is a point of convergence for the undertaking. At the end of the day, except if
high velocity is a necessity, the advantages presented by Python are sufficient to divert us from
its speed restrictions.

2. Powerless in Versatile Figuring and Programs

While it fills in as a brilliant server-side language, Python is a lot of seldom seen on the
clientside. Other than that, it is seldom at any point used to carry out cell phone based
applications.

One such application is called Carbonnelle.

The explanation it isn't the case renowned regardless of the presence of Brython is that

it isn't so secure.

3. Plan Limitations

As you probably are aware, Python is progressively composed. This implies that you
don't have to announce the kind of factor while composing the code. It utilizes duck-
composing. In any case, stand by, what's that? Indeed, it simply intends that on the off chance
that it seems to be a duck, it should be a duck.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 15
 4. Immature Information base Access Layers

Contrasted with additional generally utilized innovations like JDBC (Java Information
base Network) and ODBC (Open Data set Availability), Python's data set admittance layers
are a piece immature. Therefore, it is less normal applied in tremendous endeavors.

5. Straightforward

No, we're dead serious. Python's straightforwardness can for sure be an issue. Take my
model. I don't do Java, I'm all the more a Python individual. As far as I might be concerned,
its grammar is easy to such an extent that the verbosity of Java code appears to be pointless.
This was about the Benefits and Burdens of Python Programming Language.

7.1.2 History of Python : -

What do the letter set and the programming language Python share for all intents and
purpose? Right, both beginning with ABC. Assuming we are discussing ABC in the Python
setting, obviously the programming language ABC is implied. ABC is a broadly useful
programming language and programming climate, which had been created in the Netherlands,
Amsterdam, at the CWI (Centrum Wiskunde &Informatica). The best accomplishment of ABC
was to impact the plan of [Link] was conceptualized in the last part of the 1980s.
Guido van Rossum worked that time in a venture at the CWI, called One-celled critter, a
disseminated working framework. In a meeting with Bill Venners1, Guido van Rossum said:
"In the mid 1980s, I functioned as an implementer in a group fabricating a language called
ABC at Centrum voor Wiskunde en Informatica (CWI). I don't have the foggiest idea how
well individuals know ABC's effect on Python. I attempt to make reference to ABC's impact
since I'm obligated to all that I got the hang of during that venture and to individuals who
worked on it."Later on in a similar Meeting, Guido van Rossum proceeded: "I recalled all my
experience and a portion of my disappointment with ABC. I chose to attempt to plan a basic
prearranging language that had a portion of ABC's better properties, however without its
concerns. So I began composing. I made a basic virtual machine, a straightforward parser, and
a straightforward runtime. I made my own adaptation of the different ABC parts that I
preferred. I made an essential sentence structure, involved space for explanation gathering
rather than wavy supports or start end obstructs, and fostered few strong information types: a
hash table (or word reference, as we call it), a rundown, strings, and numbers."

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 16
 7.2 WHAT IS MACHINE LEARNING : -
Before we investigate the subtleties of different AI techniques, how about we start by
seeing what AI is, and what it isn't. AI is frequently classified as a subfield of computerized
reasoning, yet I find that classification can frequently be deluding at first brush. The
investigation of AI positively emerged from research in this specific situation, yet in the
information science utilization of AI techniques, it's more useful to consider AI for of building
models of information.

In a general sense, AI includes building numerical models to assist with grasping

information. "Learning" enters the fight when we give these models tunable boundaries that
can be adjusted to noticed information; in this way the program can be viewed as "learning"
from the information. When these models have been fit to recently seen information, they can
be utilized to foresee and grasp parts of recently noticed information. I'll pass on to the peruser
the more philosophical deviation in regards to the degree to which this kind of numerical,
model-based "learning" is like the "learning" displayed by the human [Link] the
issue setting in AI is fundamental for utilizing these apparatuses really, thus we will begin for
certain general arrangements of the sorts of approaches we'll examine here.

7.2.1 Categories Of Machine Leaning :-

At the most key level, AI can be sorted into two fundamental sorts: regulated learning

and solo learning.

Managed learning includes some way or another displaying the connection between
estimated highlights of information and some mark related with the information; when this
not entirely set in stone, applying names to new, obscure data can be utilized. This is
additionally partitioned into arrangement undertakings and relapse assignments: in order, the
marks are discrete classes, while in relapse, the names are nonstop amounts. We will see
instances of the two sorts of directed learning in the accompanying area.

Individuals, right now, are the most insightful and high level species on earth since they can
think, assess and take care of mind boggling issues.

On the opposite side, simulated intelligence is still in its underlying stage and haven't
outperformed human knowledge in numerous viewpoints. Then, at that point, the inquiry that
is the need to make machine learn? The most appropriate justification for doing this is, "to
simply decide, in light of information, with productivity and scale".

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 17
7.2.2 Challenges in Machines Learning :-

While AI is quickly developing, taking huge steps with network protection and
independent vehicles, this fragment of simulated intelligence as entire actually has far to go.
The purpose for is that ML has not had the option to beat number of difficulties. The
difficulties that ML is confronting as of now are −

Nature of information − Having great quality information for ML calculations is perhaps of

the greatest test. Utilization of bad quality information prompts the issues connected with
information preprocessing and include extraction.

Tedious assignment − One more test looked by ML models is the utilization of time

particularly for information securing, include extraction and recovery.

Absence of expert people − As ML innovation is still in its outset stage, accessibility of master

assets is a difficult situation.

No reasonable target for figuring out business issues − Having no unmistakable goal and
obvious objective for business issues is one more key test for ML since this innovation isn't
unreasonably adult yet.

Issue of overfitting and underfitting − On the off chance that the model is overfitting or

underfitting, it can't be addressed well for the issue.

Revile of dimensionality − Another test ML model appearances is an excessive number of

highlights of important pieces of information. This can be a genuine block.

Trouble in organization − Intricacy of the ML model makes it very challenging to be

conveyed, in actuality

7.2.3 Applications of Machines Learning :-

AI is the most quickly developing innovation and as per analysts we are in the brilliant year
of computer based intelligence and ML. It is utilized to take care of some certifiable complex
issues which can't be tackled with customary methodology. Following are a few genuine uses
of ML-

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 18
• Feeling examination

• Feeling investigation

• Mistake identification and avoidance

• Weather conditions guaging and expectation

• Financial exchange investigation and gauging

• Discourse blend

• Discourse acknowledgment

• Client division

• Object acknowledgment

• Extortion identification

• Extortion avoidance

• Proposal of items to client in web based shopping How to Start Learning Machine

Learning?

Arthur Samuel begat the expression "AI" in 1959 and characterized it as a "Field of study

that gives PCs the capacity to learn without being unequivocally modified".

Also, that was the start of AI! In current times, AI is one of the most famous (if not the
most!) profession decisions. As per For sure, AI Designer Is The Best Occupation of 2019
with a 344% development and a typical base compensation of $146,085 each year.

Advantages of Machine learning :-

1. Effectively recognizes patterns and examples

2. No human intercession required (robotization)
3. Taking care of complex and multi-assortment information
4. Wide Applications

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 19
 7.3 PYTHON DEVELOPMENT STEPS : -
VGuido Van Rossum distributed the primary variant of Python code (form 0.9.0) at
[Link] in February 1991. This delivery included as of now special case taking care of,
capabilities, and the center information kinds of rundown, dict, str and others. It was
likewise object situated and had a module framework.
• Print is currently a capability
• Sees and iterators rather than records
• The guidelines for requesting examinations have been rearranged. For example a
heterogeneous rundown can't be arranged, in light of the fact that every one of the
components of a rundown should be practically identical to one another.

• There is just a single number sort left, for example int. long is int too.

• The division of two whole numbers returns a float rather than a whole number. "//" can be
utilized to have the "old" conduct.

• Text Versus Information Rather than Unicode Versus 8-digit

Purpose :-

We showed that our methodology empowers fruitful division of intra-retinal layers —

even with bad quality pictures containing spot commotion, low difference, and different
power ranges all through — with the help of the ANIS include.

7.4 PYTHON
Python is a deciphered significant level programming language for broadly useful
programming. Made by Guido van Rossum and first delivered in 1991, Python has a plan
reasoning that underlines code meaningfulness, prominently utilizing critical whitespace.

Python includes a powerful kind framework and programmed memory the board. It upholds
various programming ideal models, including objectsituated, basic, practical and procedural,
and has a huge and thorough standard library.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 20
 7.5 MODULES UTILIZED IN UNDERTAKING :-

Tensorflow
TensorFlow is a free and open-source programming library for dataflow and
differentiable programming across a scope of errands. It is a representative numerical
library, and is likewise utilized for AI applications like brain organizations. It is utilized
for both examination and creation at Google.

TensorFlow was created by the Google Mind group for inside Google use. It was

delivered under the Apache 2.0 open-source permit on November 9, 2015.

Numpy

Numpy is a universally useful exhibit handling bundle. It gives an elite exhibition

multifaceted cluster article, and instruments for working with these clusters.

It is the basic bundle for logical registering with Python. It contains different elements

including these significant ones:

A strong N-layered exhibit object

Refined (telecom) capabilities

Instruments for incorporating C/C++ and Fortran code

Helpful direct polynomial math, Fourier change, and arbitrary number capacities

Other than its undeniable logical purposes, Numpy can likewise be utilized as a
proficient complex compartment of nonexclusive information. Inconsistent information
types can be characterized utilizing Numpy which permits Numpy to consistently and
rapidly coordinate with a wide assortment of data sets.

Pandas

Pandas is an open-source Python Library giving elite execution information control

and investigation instrument utilizing its strong information structures. Python was
significantly utilized for information munging and readiness. It had almost no
commitment towards information examination. Pandas tackled this issue. information, no
matter what the beginning of information load, plan, control, model, and dissect.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 21
 7.6 INSTALL PYTHON STEP-BY-STEP IN WINDOWS AND MAC :
Python a flexible programming language doesn't come pre-introduced on your PC
gadgets. Python was first delivered in the year 1991 and until now it is an exceptionally
famous undeniable level programming language. Its style reasoning stresses code
lucidness with its remarkable utilization of extraordinary whitespace.

The article arranged approach and language build given by Python empowers
developers to compose both clear and intelligent code for projects. This product doesn't
come prebundled with Windows.

Note: The python adaptation 3.7.4 can't be utilized on Windows XP or prior gadgets.

Download the Correct version into the system.

Stage 1: Go to the authority webpage to download and introduce python utilizing Google

Chrome or some other internet browser. Or on the other hand

Snap on the accompanying connection: [Link]

Fig: 7.6.1 opening the page of python

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 22
 Step 2: Snap on the Download Tab.

Fig: 7.6.2 Snap on the download tab

Presently, check for the most recent and the right variant for your working situation.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 23
 Step 3: You can either choose the Download Python for windows 3.7.4 button in Yellow
Tone or you can scroll further down and tap on download with separate to their rendition.

Here, we are downloading the latest python adaptation for windows 3.7.4

Fig: 7.6.3 Here Downloading the latest python 3.7.4

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 24
 Step 4: Look down the page until you track down the Documents choice.

Step 5: Here you see an alternate rendition of python alongside the working framework.

Fig:7.6.4 Opening several Files

• To download Windows 32-digit python, you can choose any one from the three choices:
Windows x86 embeddable compress document, Windows x86 executable installer or
Windows x86 online installer.

•To download Windows 64-digit python, you can choose any one from the three choices:
Windows x86-64 embeddable compress document, Windows x86-64 executable installer
or Windows x86-64 online installer.

Here we will introduce Windows x86-64 online installer. Here your initial segment with
respect to which adaptation of python is to be downloaded is finished. Presently we push
forward with the subsequent part in introducing python for example Establishment

Note: To know the progressions or updates that are made in the form you can tap on the

Delivery Note Choice.

Installation of Python

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 25
 Step 1: Go to Download and Open the downloaded python rendition to complete the

establishment cycle.

Fig:7.6.5 Opening the downloaded python

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 26
 Step 2: Before you click on Introduce Now, Make a point to put a tick on Add Python 3.7 to

Way.

Fig: 7.6.6 Installation of python

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 27
 Step 3: Click on Introduce NOW After the establishment is fruitful. Click on Close.

Fig: 7.6.7 Successfully downloaded python

With these over three stages on python establishment, you have effectively and accurately

introduced Python. This is the ideal opportunity to check the establishment.

Note: The establishment interaction could require two or three minutes.

Verify the Python Installation

Stage 1 : Snap on Start

Stage 2: In the Windows Run Order, type "cmd".

Stage 3: Open the Order brief choice.

Stage 4: Let us test whether the python is accurately introduced. Type python - V and press

Enter.

Stage 5: You will find the solution as 3.7.4

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 28
Check how the Python Inactive functions
Stage 1: Snap on Start

Stage 2: In the Windows Run order, type "python inactive".

Stage 3: Snap on Inactive (Python 3.7 64-bit) and send off the program
Stage 4: To proceed working Out of gear you should initially save the document. Click on

Document > Snap on Save

Fig: 7.6. 8 Saving the Code

Stage 5: Name the document and save as type ought to be Python records. Click on SAVE.

Here I have named the records as Hello World.

Stage 6: Presently for example enter print

CODE: create database Flood; use Flood; create table

signup(username varchar(50) primary key,
password varchar(50),

contact_no varchar(15),

email varchar(50),

address varchar(50));

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 29
 CHAPTER-8
SYSTEM TESTING

8.1 FRAMEWORK TESTING

The motivation behind testing is to find blunders. Testing is the most common way of
attempting to find each possible shortcoming or shortcoming in a work item. It gives a method
for really taking a look at the usefulness of parts, sub congregations, gatherings as well as a
completed item It is the most common way of practicing programming with the purpose of
guaranteeing that the

Programming framework meets its prerequisites and client assumptions and doesn't flop
in an unsatisfactory way. There are different sorts of test. Each test type tends to a particular
testing necessity.

KINDS OF TESTS

The reason for testing is to find blunders. Testing is the most common way of attempting
to find each possible shortcoming or shortcoming in a work item. It gives a method for really
taking a look at the usefulness of parts, sub congregations, gatherings as well as a completed
item It is the most common way of practicing programming with the plan of guaranteeing that
the Product framework meets its prerequisites and client assumptions and doesn't bomb in an
unsatisfactory way. There are different kinds of test. Each test type tends to a particular testing
prerequisite.

8.2 SORTS OF TESTS

Unit testing

Unit testing includes the plan of experiments that approve that the inward program
rationale is working appropriately, and that program inputs produce legitimate results. All
choice branches and interior code stream ought to be approved. It is the trying of individual
programming units of the application .it is finished after the fruition of a singular unit before
reconciliation. This is an underlying testing, that depends on information on its development
and is obtrusive. Unit tests perform essential tests at part level and test a particular business
interaction, application, as well as framework setup.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 30
 Co-ordination testing

Coordination tests are intended to test incorporated programming parts to decide whether
they really run as one program. Testing is occasion driven and is more worried about the
essential result of screens or fields. Incorporation tests exhibit that albeit the parts were
separately fulfillment, as shown by effectively unit testing, the mix of parts is right and
predictable. Reconciliation testing is explicitly pointed toward uncovering the issues that
emerge from the blend of parts.

Practical test

Practical tests give efficient exhibits that capabilities tried are accessible as indicated by

the business and specialized necessities, framework documentation, and client manuals.

Utilitarian testing is focused on the accompanying things:

Legitimate Information : recognized classes of substantial info should be acknowledged.

Invalid Information : recognized classes of invalid information should be dismissed.
Capabilities : recognized capabilities should be worked out.
Yield : distinguished classes of utilization yields should be workedout.
Frameworks/Techniques : connecting frameworks or strategies should be conjured.

Association and planning of useful tests is centered around necessities, key

capabilities, or unique experiments. Moreover, precise inclusion relating to recognize Business
process streams; information fields, predefined processes, and progressive cycles should
considered for test. Before practical testing is finished, extra tests are distinguished and the
powerful worth of current not entirely set in stone.

Framework Test

Framework testing guarantees that the whole incorporated programming

framework meets necessities. It tests a design to guarantee known and unsurprising outcomes.
An illustration of framework testing is the arrangement situated framework combination test.

Framework testing depends on process portrayals and streams, stressing pre-driven process
connections and combination focuses.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 31
 White Box Testing

White Box Testing is a trying wherein in which the product analyzer knows
about the internal operations, construction and language of the product, or if nothing else its
motivation. It is reason. Used to test regions can't be reached from a black box level.

Discovery Testing

Discovery Testing will be trying the product with practically no information on the
internal functions, design or language of the module being tried. Black box tests, as most
different sorts of tests, should be composed from a conclusive source record, for example,
detail or necessities report, for example, particular or prerequisites archive. It is a trying
wherein the product under test is dealt with, as a black box .you can't "see" into it. The test
gives inputs and answers yields disregarding the way that the product works.

Unit Testing

Unit testing is typically led as a feature of a consolidated code and unit test
period of the product lifecycle, in spite of the fact that it is entirely expected for coding and
unit testing to be led as two unmistakable stages.

Test procedure and approach

Field testing will be performed physically and utilitarian tests will be written

exhaustively.

Test targets

• All field passages should work appropriately.

• Pages should be actuated from the recognized connection.
• The passage screen, messages and reactions should not be postponed.
• Confirm that the passages are of the right configuration
• No copy sections ought to be permitted
• All connections ought to take the client to the right page.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 32
 8.3 TESTING
SOFTWARE TESTING
Testing

Testing is a process of executing a program with the aim off in ding error. To make ours of
software perform well it ought to be sans mistake. On the off chance that testing is done
effectively it will eliminate every one of the blunders from the software.

Types of Testing

1 White BoxTesting
2 Black BoxTesting

3 Unit testing
4 IntegrationTesting
5 AlphaTesting
6 BetaTesting
7 Performance Testing, etc

WhiteBoxTesting

Testing strategy in light of information on the interior rationale of an application's code and
incorporates tests like inclusion of code explanations, branches, ways, conditions. It is
performed by software developers.

Black BoxTesting

A technique for programming testing that checks the usefulness of an application without

Specific knowledge of the application's code/internal structure. Tests are ased on

requirement and functionality.

Unit Testing

Programming check and approval technique in which a software engineer tests in the event
that singular units of source code are good for use. It is generally led by the advancement
group.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 33
 IntegrationTesting

The stage in programming testing where individual programming modules are consolidated

and tried collectively. It is typically directed by testing groups.

Alpha Testing

Kind of testing a product item or framework led at the engineer's site. Normally it is

performed toward the end clients.

BetaTesting

Last testing prior to delivering application for business reason. It is regularly finished by

endclients or others.

PerformanceTesting

Practical testing led to assess the consistence of a framework or part with indicated

execution necessities. It is typically led by the presentation engineer.

Black BoxTesting

Blackbox testing will be trying the usefulness of an application without knowing the
subtleties of its implementation including internal program structure,data structures etc. Test
cases for black box testing are made in view of the prerequisite details. Hence, it is

Fig: 8.3.1 Black Box Testing

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 34
 When applied to AI models, discovery testing would mean testing AI models without
realizing the inside subtleties, for example, highlights of the AI model, the calculation used to
make the model and so on. The test, in any case, is to check the test result against the normal
qualities that are known in advance.

Fig: 8.3.2 Black Box Testing for Machine Learning algorithms

The above Fig.4.2 addresses the black box testing method for AI calculations.

Input Actual Predicted

Output Outp
ut
[16,6,324,0,0,0,22,0,0,0,0,0,0 0 0

[16,7,263,7,0,2,700,9,10,1153 1 1

,832,9,2]

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 35
 The model gives out the right result when various data sources are given which are
referenced in Table 4.1. In this way the program is supposed to be executed true to form or
address program

Input Actual Output Predicted Output

[16,6,324,0,0,0,22,0,0,0,0,0,0] 0 0

[16,7,263,7,0,2,700,9,10,1153,832,9,2 1 1

The model gives out the correct output when different inputs are given which are mentioned
in Table 4.1. Therefore the program is said to be executed as expected or correct program
Test Test Case Test Case Test Steps Test Test
Case
Cas Name Description Step Expected Actual Priorit
e Id Statu s Y

01 Start the Host the If it We The High High

Applicatio application doesn't cannot application
N and test if Start run the hosts

it
starts applicati success.

making sure on.

the required
software is

available
___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 36
 02 Home Check the If it We The High High

Page
deploymen doesn’t cannot application

t
environme load. access is running

n
t for the successfull y

properly applicati .
loading on.
the
application.
03 User Verify the If it We The High High
Mode working of doesn’t cannot application
the Respond use the displays the

application Freestyle Freestyle

in freestyle mode. Page
mode

04 Data Input Verify if the If it fails We The High High

application to take the cannot application

takes input input or proceed updates the

and updates store in further input to

application
The

Database

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 37
 CHAPTER-9
EXPERIMENTATION RESULTS

9.1 MODULES
Flood forecasting using machine learning (ML) involves a series of integrated modules that
process data and predict flood events. These modules are designed to analyze various factors,
such as rainfall, river discharge, and historical flood patterns, to make accurate forecasts.
Here's an overview of the key modules involved:

1. Data Collection: The first step in flood forecasting is gathering data from various sources. This
includes meteorological data (e.g., rainfall, temperature), hydrological data (e.g., river levels,
soil moisture), and satellite imagery. The data is usually collected in real-time through sensors,
weather stations, and remote sensing technologies.

2. Data Preprocessing: The raw data collected often requires preprocessing. This step involves
cleaning the data, handling missing values, normalizing, and transforming it into a format
suitable for machine learning models. Feature engineering is also performed, where relevant
attributes such as soil saturation or rainfall intensity are derived.

3. Model Training: Machine learning algorithms are trained on historical flood data to learn
patterns and correlations. Common ML models used for flood prediction include regression
models, decision trees, random forests, support vector machines (SVM), and neural networks.
The model learns how various factors like weather conditions and river flow affect the
likelihood of flooding.

4. Prediction and Forecasting: Once the model is trained, it can predict future flood events based
on current conditions. These predictions are made by feeding real-time data into the trained
model. The output includes flood probability, flood extent, and potential flood timing.

5. Post-processing and Decision Support: The final step involves refining the model's outputs to
provide actionable insights for disaster management. Decision support tools analyze the model's
predictions to determine areas at risk, enabling timely warnings and evacuation plans.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 38
9.2 SAMPLE SOURCE CODE
Here's a simple example of using machine learning for flood forecasting with Python and scikit-
learn. In this example, we'll use a dataset (hypothetical) with features like rainfall and river levels to
predict whether a flood will occur (binary classification).

# Import necessary libraries

import pandas as pd

from sklearn.model_selection import train_test_split

from [Link] import RandomForestClassifier

from [Link] import accuracy_score

# Load the dataset

data = pd.read_csv('flood_data.csv')

# Assume dataset has columns: 'rainfall', 'river_level', 'soil_moisture', 'flood_occurred'

X = data[['rainfall', 'river_level', 'soil_moisture']] # Features

y = data['flood_occurred'] # Target variable (1 for flood, 0 for no flood)

# Split the data into training and testing sets

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Initialize and train the Random Forest model

model = RandomForestClassifier(n_estimators=100, random_state=42)

[Link](X_train, y_train)

# Make predictions on the test set

y_pred = [Link](X_test)

# Evaluate the model's accuracy

accuracy = accuracy_score(y_test, y_pred)

print(f'Accuracy: {accuracy * 100:.2f}%')
___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 39
9.3 EXPERIMENTATION RESULTS

In this undertaking we are utilizing different AI calculation to foresee or conjecture

flood circumstance as this is a characteristic catastrophe which can cause immense loss of
lives and monetary resources. Opportune and exact forecast of future floods can help in
decrease such misfortune and to foresee flood precisely we have assessed execution different
AI calculations like SVM, Strategic Relapse, MLP and KNN. In all calculations MLP is giving
best execution and to carry out this task we have utilized beneath flood dataset from KAGGLE
site.

Fig: 9.3.1 Annual Rainfall Floods in Kerala

In above dataset first line contains dataset segment names and remaining columns
contains dataset values. In each column we have month to month and yearly precipitation and
in light of that we have class mark as YES (flood happen) and negative (no flood happen).

So by utilizing above dataset we will prepare all calculations and assess their
presentation with regards to exactness, accuracy, review, FSOCRE, awareness and
particularity.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 40
 Fig: 9.3.2 Annual Rainfall in Kerala

In above test information we have month to month and yearly precipitation without flood
name and when we apply this dataset on MLP calculation then it will foresee flood will happen
or not.

To execute this undertaking we have planned following modules

1) New Client Information exchange Here: utilizing this module we will permit client to

information exchange with the application

2) User Login: utilizing this module we will permit client to login to application

3) Preprocess Dataset: utilizing this module we will peruse flood dataset and afterward
eliminate missing qualities and afterward standardize dataset values and afterward split
dataset into train and test where application utilize 80% dataset for preparing and 20%
for testing

4) Run AI Calculations: utilizing this module we will prepare each of the 4 AI calculations
like SVM, Strategic Relapse, KNN and MLP and compute forecast exactness on test
information

5) Forecast Flood: utilizing this module we will transfer test information and afterward

6) MLP will anticipate flood from that test information

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 41
 Fig: 9.3.3 python DJANGO server

In above screen python DJANGO server began and presently open program and enter URL

as [Link] and press enter key to get underneath page.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 42
 Fig: 9.3.4 Flood Forecasting By Using machine learning

In above screen click on 'New Client Data trade Here' association with get under page .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 43
 Fig: 9.3.5 Inserting New User Signup screen

In above screen client is joining and afterward click on 'Submit' button to finish information
and get beneath yield .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 44
 Fig: 9.3.6 Submiting new user signup screen

In above screen information exchange process finished and presentl'ClientLogin'

connection to get beneath login screen.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 45
 Fig: 9.3.7 User Login Screen

In above screen client is login and after login will get underneath yield .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 46
 Fig: 9.3.8 Succesfully login

In above screen click on 'Preprocess Dataset' connection to load and processdataset and
get beneath yield .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 47
 Fig: 9.3.9 Flood graph

In above screen dataset handling finished and in chart x-hub addresses names as 0

(no flood) and 1 (flood) and y-pivot addresses number of records in that mark and presently close

above diagram to get beneath yield. By utilizing mark encoding handling method we have
changed over YES and NO to 0 and 1 as AI calculations acknowledge just numeric information .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 48
 Fig: 9.3.10 Subdivision of Annual Rainfall for flood forecasting

In above screen whole dataset process and stacked and presently click on 'Run AI Calculations'
connect to prepare all calculations and get underneath yield.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 49
 Fig: 9.3.11 Algorithm used for flood forecasting

In above screen in even arrangement we can find in all calculations MLP got most
elevated exactness as 100 percent and for each run this precision might differ from 95 to 100
percent. Presently calculations are prepared and presently click on 'Timberland Flood'
connection to get beneath screen.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 50
 Fig: 9.3.12 Flood prediction Screen

In above screen select and upload ‘[Link]’ file and then click on ‘Open’ and ‘Submit’ button to
load test data and get prediction output like below screen. This [Link] is available inside
‘Dataset’ folder .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 51
 Fig: 9.3.13 Result of Flood Forecast

In above screen in first section we can see the Precipitation month to month and yearly test
information and in last segment we can see expectation yield as 'Flood Might Happen' in red
tone and 'No Flood Happen' in green tone.

URL listing

Websites Data collected

[Link]
Searching of any information that will
be used in documentation.

[Link] SQL server it performing in mainly

depending on the one of the database
using.

[Link]

[Link], online dictionary,

encyclopedia and much more.

[Link] and
Any information searching
downloading.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 52
 CHAPTER -11
FUTURE ENHACEMENT

1. Improved Data Integration:

o Machine learning can integrate diverse data sources like satellite imagery,
weather forecasts, IoT sensors, and river gauges, enabling a more comprehensive
analysis of flood risks.
2. Deep Learning for Complex Patterns:
o Deep learning models, particularly neural networks, can handle large datasets
with spatial and temporal components, detecting complex patterns in rainfall, river
flow, and soil moisture, leading to more accurate flood predictions.
3. Real-Time Prediction with Reinforcement Learning:
o Reinforcement learning algorithms can allow flood forecasting models to adapt
in real-time based on new data, improving their responsiveness and accuracy as
conditions change.
4. Scenario Modeling and Decision Support:
o ML can simulate multiple flood scenarios, assessing the effectiveness of various
mitigation strategies like dam construction or land use changes, aiding in better
decision-making and flood management.
5. Adaptive Flood Models:
o Using ML, models can continuously evolve and improve by learning from new
data, ensuring they remain relevant and accurate over time, especially in areas where
flood patterns change due to climate variability.
6. Enhanced Early Warning Systems:
o ML can power early warning systems by predicting floods earlier and with
greater accuracy, allowing authorities to issue timely alerts and evacuations.
7. Automation with IoT and Drones:
o Integration of IoT sensors and AI-powered drones can collect real-time data on
flood-prone areas, such as river levels and rainfall intensity, improving the precision
of flood forecasts.
8. Localized Flood Forecasting:
o Machine learning can enable more localized and region-specific flood
forecasting, catering to the unique hydrological and meteorological conditions of
each area.

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 53
 CHAPTER-11
CONCLUSION

The most common way of streaming in values at the beginning time and end of the jth
time span are Ij and Ij+1 separately and comparing the worth of surge are Qj and Qj+1. AI
ready to learn and further develop framework in an unequivocal way. AI gives PC programs
that can approach or get or recover access information with learning it. AI capable with to
compute enormous informational index. Counterfeit Smart framework (AIS) used to prepare
information in such method for further developing flood guaging framework with happening
at a beginning phase improvement cautioning framework.

SCOPE :

The speculation capacity of the models and diminished the vulnerability of expectation.
The fourth was the utilization of extra enhancer calculations to work on the nature of AI
calculations, e.g., for better tuning the ANNs to arrive at ideal neuronal models. It is normal
that, through these four key advancements, flood expectation will observer critical upgrades
for both present moment and long haul forecasts. Most likely, the headway of these clever ML
strategies relies exceptionally upon the appropriate utilization of delicate processing
procedures in planning novel learning calculations. This reality was talked about in the paper,
and the delicate figuring strategies were presented as the principal givers in creating cross
breed ML techniques for what's to come .

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 54
 CHAPTER-12
REFERENCE

APA Style:

1. Sharma, A., Soni, V. K., & Shishodia, D. P. (2020). Flood prediction using machine learning
techniques: A review. International Journal of Engineering and Advanced Technology (IJEAT), 9(6),
492-499. [Link]

2. Rahman, M. M., Ali, A. K. M. A. S., & Rahman, A. S. M. M. (2021). Machine learning models for
flood forecasting: A comparative review. Water, 13(2), 169. [Link]

3. Zhao, H., Liu, L., & Wang, X. (2020). Flood prediction with deep learning models using long short-
term memory (LSTM). Environmental Modelling & Software, 131, 104755.
[Link]

4. Deka, B. S. I., Bandyopadhyay, S., & Saha, A. K. (2020). Hybrid model of machine learning
algorithms for flood forecasting. Journal of Hydrology, 582, 124476.
[Link]

5. Patel, N., Bedi, H. S. P. S., & Chandra, S. (2018). Real-time flood forecasting using support vector
machine and backpropagation neural network. Hydrological Processes, 32(1), 160-174.
[Link]

6. Zhang, X., Yang, X., & Li, Z. (2021). A novel hybrid model combining machine learning algorithms
for flood prediction. Journal of Hydrology, 592, 125748.
[Link]

7. Bhaskaran, M., Reddy, P. V. C., & Sivaprasad, D. (2022). Flood risk assessment and prediction using
machine learning algorithms. Environmental Science and Pollution Research, 29(25),
37421-37438. [Link]

8. Prasad, J. V. S. R., Anjaneyulu, K. S. R., & Rao, S. M. K. (2019). Flood prediction and forecasting
using artificial neural networks. Journal of Hydrology, 575, 383-392.
[Link]

9. Amarante, J. R. G., Santos, M. A., & Silva, F. S. D. (2021). Deep learning for flood prediction:
A review. Earth Science Informatics, 14(1), 105-116. [Link]

10. Singh, A., Pal, N. S., & Shukla, R. (2021). Artificial intelligence for flood prediction: A review of
machine learning models. Environmental Monitoring and Assessment, 193(8), 526.
[Link]

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 55
 MLA Style:

1. Sharma, A., V. K. Soni, and D. P. Shishodia. "Flood Prediction Using Machine Learning
Techniques: A Review." International Journal of Engineering and Advanced Technology
(IJEAT), vol. 9, no. 6, 2020, pp. 492-499. [Link]

2. Rahman, M. M., A. K. M. A. S. Ali, and A. S. M. M. Rahman. "Machine Learning Models for Flood
Forecasting: A Comparative Review." Water, vol. 13, no. 2, 2021, p. 169.
[Link]

3. Zhao, H., L. Liu, and X. Wang. "Flood Prediction with Deep Learning Models Using Long Short-Term
Memory (LSTM)." Environmental Modelling & Software, vol. 131, 2020, p. 104755.
[Link]

4. Deka, B. S. I., S. Bandyopadhyay, and A. K. Saha. "Hybrid Model of Machine Learning Algorithms for
Flood Forecasting." Journal of Hydrology, vol. 582, 2020, p. 124476.
[Link]

5. Patel, N., H. S. P. S. Bedi, and S. Chandra. "Real-Time Flood Forecasting Using Support Vector
Machine and Backpropagation Neural Network." Hydrological Processes, vol. 32, no. 1, 2018, pp. 160-
174. [Link]

6. Zhang, X., X. Yang, and Z. Li. "A Novel Hybrid Model Combining Machine Learning Algorithms for
Flood Prediction." Journal of Hydrology, vol. 592, 2021, p. 125748.
[Link]

7. Bhaskaran, M., P. V. C. Reddy, and D. Sivaprasad. "Flood Risk Assessment and Prediction Using
Machine Learning Algorithms." Environmental Science and Pollution Research, vol. 29, no. 25, 2022,
pp. 37421-37438. [Link]

8. Prasad, J. V. S. R., K. S. R. Anjaneyulu, and S. M. K. Rao. "Flood Prediction and Forecasting Using
Artificial Neural Networks." Journal of Hydrology, vol. 575, 2019, pp. 383-392.
[Link]

9. Amarante, J. R. G., M. A. Santos, and F. S. D. Silva. "Deep Learning for Flood Prediction: A Review."
Earth Science Informatics, vol. 14, no. 1, 2021, pp. 105-116.
[Link]

10. Singh, A., N. S. Pal, and R. Shukla. "Artificial Intelligence for Flood Prediction: A Review of Machine
Learning Models." Environmental Monitoring and Assessment, vol. 193, no. 8, 2021, p. 526.
[Link]

___________________________________________________________________________________
BRIL-CSE(AIML) PAGE 56