Introduction to

Machine Learning
(ML)
✓ Introduction to ML
✓ ML Vs Traditional
Programming
✓ How ML Works ?
✓ Why ML ?
✓ Types of ML
✓ ML tools

1
Introduction to Machine learning
➢ Human Learns from past experience.

➢ Machines follow instructions given by Humans.

✓ Humans Can Train The Machines

✓ To learn from past data and do what humans can do

➢ Suppose Bil is Listening to songs…

➢ Based on Tempo, intensity

• Likes
• Dislikes
ML: answer questions using data
2
3
4
What is Machine Learning?

• ML is a subset of artificial intelligence.

• It focuses on the development of computer algorithms that improve

automatically through experience and by the use of data.

• Machine Learning enables a program to learn from experiences and

improve itself at a task without being explicitly programmed.

• It concerned about designing systems algorithms that can learn from

a data with out being explicitly coded by programmer.
5
What is Machine Learning?

• It gives power to machines to execute tasks and capable of learning

and making predictions based on some experience.

• Machine learning is about extracting knowledge from data.

• Machine learning combines data with statistical tools to predict an

output.

6
ML vs traditional programing
Traditional Programming Machine Learning
• we would feed the input data and • input data is fed into the machine
a well-written and tested during the learning phase, and it
program into a machine to
generate output. works out a program for itself.
• a programmer codes all the rules • The system learns how the input
for the system to generate an and output data are correlated
output.
• When the system grows complex, and it writes a rule.
more rules need to be written & • The programmers do not need to
It can quickly become
unsustainable to maintain. write new rules each time there is
• Drawback: The programmer new data.
should code everything to • The algorithms adapt in response
distinguish between objects in
the data set. to new data and experiences to
improve efficacy over time. 7
ML vs traditional programing

Traditional Programming Machine Learning

8
Machine Learning programming

9
 How Does Machine Learning Work?
▪ ML algorithms analyze and identify patterns from datasets and use this
information to make better predictions on new datasets.

▪ The way the machine learns is similar to how humans learn and improve.
✓ Whenever we make a decision, we consider our past experiences to
assess the situation better.
✓ A machine learning model does the same by analyzing historical data to
make predictions or decisions.
✓ After all, machine learning is an AI application that enables machines to
self-learn from data. 10
How Does Machine Learning Work?
Example 1:dinosaur game - you would use the trial and error method
to learn how to play the game.
• An AI application would also learn similarly.

Example2: Sequence

3 – 9 : 4 – 16 : 5 – 25 So if you were given the number 6, which

number would you pick to match the sequence?

11
How Does Machine Learning Work?
• Inputting training data into the selected algorithm.

• Training data being known or unknown data to develop the final

Machine Learning algorithm.

• New input data is fed into the machine learning algorithm to

test whether the algorithm works correctly.

• The prediction and results are then checked against each other.

12
How Does Machine Learning Work?
• If the prediction and results don’t match, the algorithm is re-trained
multiple times until the data scientist gets the desired outcome.

13
 Machine Learning Process
Life of Machine Learning programs /Workflow

1. Define a question 6. Collect feedback

7. Refine the algorithm
2. Collect data
/Fine Tuning
3. Visualize data/Preparation
8. Loop 4-7 until the results are
4. Train algorithm
satisfying
5. Test the Algorithm/Evaluation 9. Use the model to make a
prediction
14
15
Why do we need Machine Learning?
• Machine Learning is increasingly being used to automate processes.

• Machine Learning-based applications such as smart power grid, smart

homes, smart retail, precision farming, smart cities, and urban planning.

• Machine Learning is used along with IoT to build a network of self-

learning and self-operating systems.

• Machine Learning-solutions are also used to create self-correcting

programs and automate repetitive tasks in a work process.

16
17
Why do we need Machine Learning?

Application of ML

• Finance Industry

• Government organizations

• Healthcare industry

• Education

18
19
20
21
Types of Machine Learning

22
 Supervised learning
• Used to train machine using known or labeled data for the training data. Since
the Data is known, the learning is supervised(i.e., directed into successful
execution.

• Takes labeled data input and map it to the known outputs.

✓ Supervised learning techniques need external supervision to train model

Supervised learning is analogous to training a child to walk.

✓ You will hold the child’s hand, show him how to take his foot forward, walk
yourself for a demonstration and so on, until the child learns to walk on his
own 23
24
25
26
 Cont.…
• Categories of supervised learning/Problem: Classification and
Regression

• Application: Weather Predication, Sales Forecasting and Stock Analysis

✓ Predicting real estate prices.
✓ Classifying whether bank transactions are fraudulent or not
✓ Finding disease risk factors
✓ Determining whether loan applicants are low-risk or high-risk
✓ Predicting the failure of industrial equipment's mechanical parts
27
 Un-supervised learning
• Uses unknown/unlabeled data- meaning that no one has looked at the
data before to train machines.

• Understand the pattern and trends in the data and discover the output.

• Do not need any supervisor/human intervention to train model

• Example: the unknown data consists of apples and pears which look
similar to each other. The trained model tries to put them all together
so that you get the same things in similar groups.

28
29
 Cont...
• Categories of supervised learning/Problem: Clustering and Association

• Application: Customer Segmentation and Churn Analysis

✓ Creating customer groups based on purchase behavior
✓ Grouping inventory according to sales and/or manufacturing metrics
✓ Pinpointing associations in customer data (for example, customers
who buy a specific style of handbag might be interested in a specific
style of shoe)

30
Reinforcement Learning
• The machine learns from its experience

• Uses agent and environment to produce an action and rewards.

• Follows try and error method to arrive at the desired solution.

• Example: Suppose there is a child in a room. He sees a fireplace and tries to

approach it. He feels warm and thinks that it is a good thing. But when he
tries to touch it, he gets hurt. So, he understands that fire is a bad thing
when he is too close to it otherwise it is good.
• The concept behind reinforcement learning i.e. the machine learns from
the surrounding by interacting with it. 31
Reinforcement Learning

• Markov decision process

• Q-learning
• Policy gradient
32
33
Types of Machine learning summary

34
 Questions ?

1. Explain the following Machine Learning Tools:

Microsoft Azure Machine Learning, IBM Watson,
TensorFlow, Amazon Machine Learning, OpenNN,
PyTorch, Vertex AI, BigML, Apache Mahout, Weka
2. What are the other types of machine learning tools ?

35
What is Deep Learning?
▪ Deep learning is a subset of machine learning that deals with
algorithms inspired by the structure and function of the human
brain.
▪ Deep learning algorithms can work with an enormous amount of
both structured and unstructured data.
▪ Deep learning’s core concept lies in artificial neural networks,
which enable machines to make decisions like human.
▪ Machine learning algorithms usually require structured data.
36
 Types of Algorithms used in Deep Learning
1. Convolutional Neural Networks (CNNs) – used for image data
2. Long Short Term Memory Networks (LSTMs) - used to learn,
process, and classify sequential data.
3. Recurrent Neural Networks (RNNs) – used for sequential data
4. Generative Adversarial Networks (GANs) - can be used to generate
videos, such as synthesizing new scenes in a movie.
5. Radial Basis Function Networks (RBFNs) - used for used for
function approximation problems.
37
 Types of Algorithms used in Deep Learning
6. Multilayer Perceptron (MLPs) - powerful data-driven modeling tool
in ANNs
7. Self Organizing Maps (SOMs) - widely used neural network for
dimensional reduction and clustering. can be used for exploratory data
analysis, and visualization of high dimensional datasets.
8. Deep Belief Networks (DBNs) - used for
9. Restricted Boltzmann Machines( RBMs) - used for unsupervised
learning tasks.
10.Autoencoders - used for feature selection and feature extraction 38
39
How Do RNNs and MLPs work?

How Do
RNNs
work?

How Do
MLPs
Work?

40
How Do CNNs Work?

41
42
43
Difference between Machine Learning vs Deep learning
Machine learning Deep learning

A subset of AI A subset of machine learning

Can train on smaller data sets Requires large amounts of data

Requires more human intervention to correct Learns on its own from environment and past
and learn mistakes

Shorter training and lower accuracy Longer training and higher accuracy

Makes simple, linear correlations Makes non-linear, complex correlations

Can train on a CPU (central processing unit) Needs a specialized GPU (graphics
processing unit) to train
45
46
Thank You