WEEK 1 PRACTICE ASSIGNMENT
WEEK 1 PRACTICE ASSIGNMENT
[Link] for applications of AI/ML in your area of specialisation.
Answer:Smart phones have AI in it, as they do services like digital assistants,
chatbots, social media websites, and much more. Many home electronics also
use AI, such as robot vacuum cleaners, security systems,using voice
recognition to do simple work like switching on light,ac and more. And there
are classic examples of auto-navigation and [Link] learning powers
applications like predictive analytics, image and speech recognition,
recommendation systems, and autonomous vehicles, and is widely used
across sectors like healthcare, finance, retail, and more.
[Link] what age do children learn to differentiate between a dog and a
cat? How do they acquire this skill? You would have forgotten how you
learnt it, but can you look around your apartment and observe how
children learn this distinction?
Answer:Usually children can recognise a dog and a cat between 6 months to
12 months after they are [Link] differentiate by auditory and visual skills
with the help of parental [Link] learn by observing the physical
characteristics of dogs and cats – their size, shape, fur, and sounds. They
also learn by observing how others interact with these animals.
[Link] does a typical spam message look like? Can you come up with
five examples of messages that are spam? You can either be creative
and come up with your examples or search the web. Now go back and
look at the five examples you generated. What are some patterns that
you see?
Answer:A typical spam text message often appears suspicious and lacks
genuine context.
● You have won $100000 in lottery. Please provide us your account
details.
● Download this software - your system is compromised
● Click this link for account verification.
● Your package is returned due to your absence in your home,click the
link for refunding the money
● We offer jobs please click the link to fill in your details for further call
from our company.
These messages mostly target aged person or teen or youth persons.
Mainly every spam messages for personal details.
[Link] about some of the features that will help you make an informed
decision about providing loans to customers at a bank. Do any of your
connections work at a bank? Can you ask them about the process that
is followed before sanctioning a loan? Now think about how this
process can be automated with the help of a machine.
Answer:Features that help in making an informed loan decision:
*Credit history and score
*Income and employment status
*Loan purpose and amount
*Collateral
No,any of personal connections who work at a bank does not help in
providing loans.
i)Application submission
ii)Income and employment verification
iii)Collateral assessment
iv)Past credit history
v) Risk assessment
vi)Decision making
vii)Sanction Letter Issuance
viii)Loan Disbursement
The traditional, manual loan sanctioning process is time-consuming and
can be prone to human error and bias.
*Automated Data Collection and Preprocessing
*Machine Learning for Credit and Risk Scoring
*Fraud Detection
*Automated Decisioning and Workflow
[Link] is a collection of handwritten images. A collection of images
is called an image dataset. This dataset is used to teach a machine to
distinguish between different digits. Use this resource:
([Link] to explore the
dataset and answer the following questions:
(1) How many images does the dataset have in total?
ANSWER:The images are divided into a training set of 60,000 images and a test set
of 10,000 images.
(2) Note down your observations about the images corresponding to a
single digit, say 0.
ANSWER:i)Even though the all digits are same,there is a wide variety of
styles.
ii)The digit 0 shows a high degree of variability in its handwritten form
compared to other digits like the number 1.
iii)Most images of the digit 0 are oval or circular in shape
iv)Some images are blurry so that it will be a little difficult to identify for
both humans and machine.
(3) Look at the images corresponding to the digit 5. Specifically, look at
sample index 2258 under the digit 5. What number does it actually look
like?
ANSWER:Sample index 2258 under the digit 5 looks like 6.
(4) Write down the ten digits and identify similarities and differences in
the strokes that you make. Are there some digits that share similar
strokes, say 8 and 0? Can you identify similar families of digits?
ANSWER: 0,1,2,3,4,5,6,7,8,9
Similarities:
● Use of curved and straight lines: All digits are formed using a
combination of straight and curved lines.
● Direction of flow: Many digits are typically written starting from the top
and moving downwards (e.g., 1, 4, 7, 9).
Differences:
● Number of strokes: Some digits can be written with a single, continuous
stroke (e.g., 0, 8), while others require multiple separate strokes (e.g., 4,
5, 7).
● Closed loops: Digits 0, 6, 8, and 9 all contain at least one closed loop,
while the others do not.
(5) What are some of the challenges in teaching a machine to identify
digits?
ANSWER:One of the most common challenges in teaching a machine to
identify digits is the sheer variability and unpredictability of human
handwriting. Even with a seemingly simple task like recognizing digits from 0
to 9, machines face several hurdles that humans overcome instinctively.
(6) Where do you think such an application would be useful?
ANSWER:i)Postal Services
ii)Data Entry and Digitization
iii)Banking and Finance
[Link] the height and weight of students in your class and present it
in the form of a table. You can use Google sheets for this. Do not
include the names of the students. The table should have just two
columns, height and weight. Plot a scatter plot of this data. Do you
observe any patterns?
As height increases, there's a possibility for weight also to increase. The data
points, while not forming a perfect straight line,they go upward from left to
right. This indicates that taller individuals in this dataset typically have a
higher weight.
BY:[Link]