Unit 1 : AI Project Cycle

Short Question Answers:-

1. What stages of an AI project are there?
Ans:- Problem Scoping, Data Acquisition, Data Exploration, Modelling, Evaluation and
Deployment. Fig. Pg. 129
2. What is problem scoping?
Ans:- Problem scoping is the first phase of the AI project cycle. In this phase of AI development,
problems are identified. This is followed by design, development or construction and final project
testing. The wrong scope of the problem also leads to failure. Identifying such a problem and
having the vision to solve it is what problem scoping is all about.
3. Who are stakeholders?
Ans:- Stakeholders are those who are affected by the issue and who stand to gain from the
solution.
4. What is Rule based Approach?
Ans:- Rule Based Approach Refers to the AI modelling where the relationship or patterns in data
are defined by the developer. The machine follows the rules or instructions mentioned by the
developer, and performs its task accordingly.
5. What is Data Modelling?
Ans:- Data Modelling is all about creating systems to organize and analyze data. Think of it as
building a framework that helps AI spot patterns, make predictions, and continuously learn and
improve.
6. What are data features?
Ans:- Data features are individual measurable properties or attributes of the data that are used
as input variables in data analysis or machine learning models. They represent specific aspects
of the data, such as age, income, or temperature, which help algorithms identify patterns, make
predictions from the dataset. Good feature selection improves model performance.
7. What is data exploration?
Ans:- Data exploration is the third step of data analysis, used to examine and visualize data to
gain early insights or identify areas or patterns to dig deeper. By using interactive dashboards
and point-and-click data exploration, users can better understand the big picture and gain insights
more quickly.

Class 10/ Ch-1/ AI Project Cycle/ 1

C
Long Question Answers:-
1. How can you figure out the data using problem scoping?
Ans:- Problem scoping helps determine the necessary data by first clarifying the problem,
identifying key assumptions, and defining the scope of analysis. This process involves
understanding the problem's domain, parameters, and the specific data needed to address it.
i. Clarify the Problem Statement: Start with a well-defined problem statement that outlines
the issue you're trying to solve.
ii. Identify Key Assumptions: Consider what you're assuming about the problem and its
context. These assumptions can influence the data you need.
iii. Define the Scope of Analysis: Determine the boundaries of your analysis. This includes
the timeframe, geographical area, and specific aspects of the problem you'll focus on.
iv. Understand the Problem Domain: Gain insight into the industry or field where the
problem exists. This helps to understand the challenges, and intricacies of that domain, which
can influence the types of data needed.
v. Identify Data Requirements: Based on the clarified problem, scope, and domain, identify
the specific data points and types needed to address the problem. This might include
quantitative data, qualitative data, or a combination of both.
2. What is 4Ws Problem Canvas?
Ans:- The 4Ws problem canvas is the basic template while scoping a problem and using this
canvas, the picture becomes clearer while we are working to solve it.
a) Who: The “Who” block helps you in analyzing the people getting affected directly or indirectly
due to it? Under this, you find out who the ‘stakeholders’ to this problem are and what you know
about them. Stakeholders are the people who face this problem and would be benefitted with the
solution.
b) What: Under the “What” block, you need to look into what you have on hand. At this stage,
you need to determine the nature of the problem. What is the problem and how do you know that
it is a problem?
c) Where: In this block, you need to focus on the context/situation/location of the problem. It will
help you look into the situation in which the problem arises, the context of it, and the locations
where it is prominent.
d) Why: In the “Why” canvas, think about the benefits which the stakeholders would get from the
solution and how would it benefit them as well as the society.
3. What is data acquisition? How many types of data acquisition methods are there?
Ans:- It is a process of acquiring data from various sources. Data can be collected through
surveys, interviews, web scraping, sensors, cameras and observations. Data is collected before

Class 10/ Ch-1/ AI Project Cycle/ 2

C
data exploration. The data which is fed into the model is training data and prediction data is testing
data. There are mainly 2 types of data collection methods:-
i. Primary Data collection method:- The word “primary data” describes data that you
independently gather as opposed to data that you compile after being originally recorded by
another party. First party or primary data is information that has been collected directly from
the source. First party data is very accurate and relevant to your business because it is
collected directly from your target audience. This information is extremely valuable because it
offers perceptions into the attitudes, preferences and traits of your clients.
ii. Secondary Data collection method:- Second party secondary data can be acquired directly
from the company that gathered it or through a private market. Many of the advantages of first
party data also apply to second party data. You can be sure of its correctness because it
comes directly from the source, but it also provides you with information that you wouldn’t
otherwise have access to. Furthermore, second party data offers you insights that you might
not have been able to get from only your first party data.
4. What is the difference between Training Data & Testing Data?
Ans:-
Training Data Testing data
The portion of the original data used to train Testing data is used to evaluate the model’s
the machine learning model is known as accuracy.
Training Data.
The size of the training dataset is typically The testing dataset is typically not very large.
greater.
The model is familiar with training data since it For the model, test data is equivalent to
uses it to train the model. fresh or new information.
Training data set is the one on which we train Test data is used only to assess performance
and fit our model basically to fit the of model.
Parameters.
Training data's output is available to model. Testing data is the unseen data for which
predictions have to be made.

5. What are the different types of Learning based approach?

Ans:- The main types of learning-based approaches are supervised learning, unsupervised
learning, and reinforcement learning.
Supervised learning is a learning in which we teach or train the machine using data which is
well labelled that means some data is already tagged with the correct answer. After that, the
machine is provided with a new set of examples (data) so that supervised learning algorithm
analyses the training data (set of training examples) and produces a correct outcome from
labelled data.

Class 10/ Ch-1/ AI Project Cycle/ 3

C
An unsupervised learning approach works on an unlabeled dataset. This means that the data
which is fed to the machine is random and there is no knowhow available about it to the trainer.
These learning models are used to identify trend, pattern and relationship in the data which is fed
into it. In this model the major features are identified by the machine, which helps the user in
understanding the data.
Reinforcement Learning: This approach involves training a model to make decisions in an
environment to maximize a reward. The model learns through trial and error, receiving rewards
or penalties for its actions. A classic example is training a model to play a video game, where it
learns to maximize its score by exploring different actions.
6. What is Evaluation? Describe the process involved in it.
Ans:- Evaluation is a very important stage of AI Project designing and training where we properly
test the system to find out the efficiency and performance of the model. After the model is
designed and trained then the reliability of the model is checked using Testing Data acquired at
the Data Acquisition Stage. This testing data is given as an input to the newly created AI model
and the output received is checked and evaluated on the basis of :
Accuracy Precision Recall F1 Score
7. Give a brief introduction to the Turing Test in AI. What is the significance of Turing Test
also write its limitations.
Ans:- The Turing Test, proposed by Alan Turing, is a method to determine if a machine can
exhibit behavior indistinguishable from that of a human. It involves a human judge interacting with
both a machine and a human through text, and the judge tries to identify which respondent is the
machine. If the judge cannot consistently identify the machine, the machine is considered to have
passed the test.
Significance: The Turing Test has been a significant benchmark in the field of AI, serving as a
test of whether a machine can exhibit human-like intelligence through natural language
interaction. It raises important questions about the nature of intelligence and how we define
thinking.
Limitations: It's important to note that passing the Turing Test does not necessarily mean that a
machine truly understands or "thinks" like a human. The machine may simply be skilled at
mimicking human behavior and responses without having any underlying understanding or
consciousness.
8. Why shouldn’t the training data be used for evaluation?
Ans:- Training data shouldn't be used for evaluation because it can lead to biased and overly
optimistic results. When a model is evaluated on the same data it was trained on, it may appear
to perform very well—not because it's genuinely good at generalizing, but because it has
memorized the patterns in that data. This is known as overfitting.

Class 10/ Ch-1/ AI Project Cycle/ 4

C
To accurately measure how well a model will perform on unseen data, it should be tested on a
separate validation or test set. This ensures the evaluation reflects the model’s true ability to
generalize to new, real-world inputs. Using training data for evaluation defeats this purpose and
can lead to poor performance when the model is deployed.
Extra Questions:
1. What are Sustainable Development Goals?
Ans:- When we cannot observe a problem around us then we should refer to the 17 goals that
have been announced by the United Nations as the Sustainable Development Goals. These goals
are to be achieved by 2030 as pledged by member nations of the UN. Artificial Intelligence
supported solutions are suggested to assist the society and government to achieve these goals
that would work to improve the lives of the people living in the society all across the nations.
2. Five sustainable Development Goals are mentioned below. Write 2 problems under
each goal that you think should be addressed for achieving the goal.
a. Quality Education b. Reduced Inequalities c. Life on Land
d. No Poverty e. Clean Water and Sanitation
Ans:- a. Quality Education:
i. Providing education remotely, using hi-tech, low-tech and no-tech approaches.
ii. Ensure equitable quality education for all.
b. Reduced inequalities:
i. Reduction of relative economic inequalities in some countries having poorest and most
vulnerable communities.
ii. Improving the situations in countries with weaker health systems.
c. Life on Land:
i. Prevention of Deforestation caused by humans and restoration of land.
ii. Preventions and cure of diseases that are transmissible between animals and humans.
d. No Poverty
i. Creation of Strong social protection systems to prevent people from falling into poverty.
ii. Reduction of social exclusion and high exposure of certain populations to disasters and
diseases.
iii. Responsible distribution of resources.
e. Clean Water and Sanitation
i. To increase access to clean drinking water and sanitation mostly in rural areas
ii. Managing our water sustainably, to manage our production of food and energy.
3. What is a problem statement template and what is its significance?
Ans:- To define and explain a problem that needs to be fixed, use the problem statement form.

Class 10/ Ch-1/ AI Project Cycle/ 5

C
Problem Statement Template helps us to summarize all the key points into one single Template
so that in future, whenever there is a need to look back on the basis of the problem, we can take
a look at the Problem Statement Template and understand the key elements of it.
4. What precautions to be taken while acquiring data for developing an AI Project?
Ans:- Data should be collected from an authentic source and should be accurate. The redundant
and irrelevant data should not be a part of prediction.
5. What are the two different approaches for AI modelling? Define them.
Ans:- Rule Based Approach Refers to the AI modelling where the relationship or patterns in data
are defined by the developer. The machine follows the rules or instructions mentioned by the
developer, and performs its task accordingly. Whereas in Learning based approach, the
relationship or patterns in data are not defined by the developer. In this approach, random data
is fed to the machine and it is left to the machine to figure out patterns and trends out of it
6. What is the purpose of getting AI Ready?
Ans:- The world is changing with each day and we have huge data coming our way. The purpose
of getting AI ready means taking steps to collect data around relevant systems, equipment, and
procedures; and storing and curating that data in a way that makes it easily accessible to others
for use in future AI applications.
7. What are the different types of sources of data from where we can collect reliable and
authentic datasets? Explain in brief.
Ans:- Data can be a piece of information or facts and statistics collected together for reference
or analysis. Whenever we want an AI project to be able to predict an output, we need to train it
first using data There could be many ways and sources from where we can collect reliable and
authentic datasets namely Surveys, Web scrapping, Sensors, Cameras, Observations,
Research, Investigation, API etc.
Sometimes Internet is also used to acquire data but the most important point to keep in mind is
that the data should be taken from reliable and authentic websites only. Some reliable data
sources are UN, Google scholar, Finance, CIA, [Link] etc.
8. Differentiate between Rule based Approach and Learning based Approach.
Rule based Approach Learning based Approach
The machine follows the rules or Instructions The machine learns by itself. It gets trained on
mentioned by the developer and performs its the data fed to it. Also known as machine
task. learning.
Rule based systems are simple kind of AI Learning based system’s principle is to
which uses a series of IF- THEN statements simulate intelligence. The ability to learn cause
to reach to a conclusion. adaptive intelligence i.e new knowledge is
acquired from existing knowledge.
Example is Decision Tree Example is Neural Network

Class 10/ Ch-1/ AI Project Cycle/ 6

C
9. What is Data Visualisation? Mention two data visualisation tools.
Ans. Data visualization is the graphical representation of information and data. By using visual
elements like charts, graphs, and maps, data visualization tools provide an accessible way to
see and understand trends and patterns in data. Two popular data visualization tools are Tableau
and Google Facets.
Tableau: It helps users visualize data in interactive dashboards and reports. It connects to
various data sources like Excel, SQL databases, Google Sheets etc. It allows drag and drop
functionality to create charts, maps and graphs. It is used by Data scientists and Analysts.
Google Facets: It is very useful visualization tool that works much faster without requiring a lot
of coding. This can be used for data visualization to interpret and analyze learning data. Facet
provides an overview layer to get summary statistics on each dataset.
10. Explain Deployment stage of AI project cycle.
Ans: Model deployment involves placing a fully operational machine learning model in a real-
world environment to generate predictions that aid decision-making. For data scientists, this is
the ultimate goal, a s it allows the model to fulfill its purpose. Deployment is a critical step in the
machine learning workflow and often presents the biggest challenge, requiring significant time
and resources. Not all models reach this stage if they fail to meet criteria.

Class 10/ Ch-1/ AI Project Cycle/ 7

C