Machine Learning

• Machine learning is a growing technology which

Machine Learning enables computers to learn automatically from past
data. Machine learning uses various algorithms for
building mathematical models and making
predictions using historical data or information.
Dr. Rabins Porwal
• Currently, it is being used for various tasks such as
image recognition, speech recognition, email
filtering, Facebook auto-tagging, recommender
system, and many more.

1 2

What is Machine Learning Contd…

Contd…
• In the real world, we are • Machine Learning is said as a subset of artificial intelligence that is
surrounded by humans who can mainly concerned with the development of algorithms which allow a
learn everything from their computer to learn from the data and past experiences on their own.
experiences with their learning • The term machine learning was first introduced by Arthur Samuel in
capability, and we have 1959. It can be summarized as:
computers or machines which “Machine learning enables a machine to automatically learn from data,
work on our instructions. But can improve performance from experiences, and predict things without being
a machine also learn from explicitly programmed.”
experiences or past data like a
human does? So here comes the
role of Machine Learning.

3 4

Contd…
Contd… How does Machine Learning work
• With the help of sample historical data, which is known as training • A Machine Learning system learns from historical data, builds the
data, machine learning algorithms build a mathematical model that prediction models, and whenever it receives new data, predicts the
helps in making predictions or decisions without being explicitly output for it.
programmed. • The accuracy of predicted output depends upon the amount of data, as the
huge amount of data helps to build a better model which predicts the output
• Machine learning brings computer science and statistics together for more accurately.
creating predictive models. Machine learning constructs or uses the
algorithms that learn from historical data. The more we will provide • Suppose we have a complex problem, where we need to perform some
the information, the higher will be the performance. predictions, so instead of writing a code for it, we just need to feed the
data to generic algorithms, and with the help of these algorithms,
• A machine has the ability to learn if it can improve its performance by machine builds the logic as per the data and predict the output.
gaining more data.

5 6

Prepared by: Dr. Rabins Porwal 1

 Contd…
Contd… Features of Machine Learning:
• Machine learning has changed our way of thinking about the problem. • Machine learning uses data to detect various patterns in a given
dataset.
• It can learn from past data and improve automatically.
• It is a data-driven technology.
• Machine learning is much similar to data mining as it also deals with
the huge amount of the data.

7 8

Need for Machine Learning Importance of Machine Learning

• The need for machine learning is increasing day by day. • The importance of machine learning can be easily understood by its uses.
• The reason behind the need for machine learning is that it is capable of doing • Currently, machine learning is used in self-driving cars, cyber fraud detection, face
tasks that are too complex for a person to implement directly. As a human, we recognition, and friend suggestion by Facebook, etc.
have some limitations as we cannot access the huge amount of data
manually, so for this, we need some computer systems and here comes the • Various top companies such as Netflix and Amazon have built machine
machine learning to make things easy for us. learning models that are using a vast amount of data to analyse the user
• We can train machine learning algorithms by providing them the huge interest and recommend product accordingly.
amount of data and let them explore the data, construct the models, • Following are some key points which show the importance of Machine
and predict the required output automatically. Learning:
• The performance of the machine learning algorithm depends on the • Rapid increment in the production of data
amount of data, and it can be determined by the cost function. • Solving complex problems, which are difficult for a human
• With the help of machine learning, we can save both time and money. • Decision making in various sector including finance
• Finding hidden patterns and extracting useful information from data.

9 10

Classification of Machine Learning 1. Supervised Learning

• Supervised learning is a type of machine learning method in which we provide
• At a broad level, machine sample labeled data to the machine learning system in order to train it, and on
learning can be classified into that basis, it predicts the output.
three types: • The system creates a model using labeled data to understand the datasets and
1. Supervised learning learn about each data, once the training and processing are done then we test
the model by providing a sample data to check whether it is predicting the exact
2. Unsupervised learning output or not.
3. Reinforcement learning • The goal of supervised learning is to map input data with the output data. The
supervised learning is based on supervision, and it is the same as when a
student learns things in the supervision of the teacher.
• Example: spam filtering
• Supervised learning can be grouped further in two categories of algorithms:
• Regression
• Classification

11 12

Prepared by: Dr. Rabins Porwal 2

 2. Unsupervised Learning 3. Reinforcement Learning
• Unsupervised learning is a learning method in which a machine learns • Reinforcement learning is a feedback-based learning method, in
without any supervision. which a learning agent gets a reward for each right action and gets a
• The training is provided to the machine with the set of data that has not penalty for each wrong action.
been labeled, classified, or categorized, and the algorithm needs to act • The agent learns automatically with these feedbacks and improves its
on that data without any supervision. performance.
• The goal of unsupervised learning is to restructure the input data into new • In reinforcement learning, the agent interacts with the environment and
features or a group of objects with similar patterns. explores it.
• In unsupervised learning, we don't have a predetermined result. The • The goal of an agent is to get the most reward points, and hence, it improves
machine tries to find useful insights from the huge amount of data. its performance.
• It can be further classifieds into two categories of algorithms: • The robotic dog, which automatically learns the movement of his
• Clustering arms, is an example of Reinforcement learning.
• Association

13 14

History of Machine Learning The early history of ML (Pre-

(Pre-1940):
• Before some years (about 40-50 • 1834: In 1834, Charles Babbage, the father of the computer,
years), machine learning was conceived a device that could be programmed with punch cards.
science fiction, but today it is the However, the machine was never built, but all modern computers rely
part of our daily life. Machine on its logical structure.
learning is making our day-to-day
life easy from self-driving cars to • 1936: In 1936, Alan Turing gave a theory that how a machine can
Amazon virtual assistant "Alexa". determine and execute a set of instructions.
• However, the idea behind machine
learning is so old and has a long
history.
Milestones which have occurred in the history of machine
learning

15 16

The era of stored program computers: Computer machinery and intelligence:

• 1940: In 1940, the first manually operated computer, "ENIAC" was • 1950: In 1950, Alan Turing published a seminal paper, "Computer
invented, which was the first electronic general-purpose computer. Machinery and Intelligence," on the topic of artificial intelligence. In
After that stored program computer such as EDSAC in 1949 and his paper, he asked, "Can machines think?"
EDVAC in 1951 were invented.
• 1943: In 1943, a human neural network was modeled with an
electrical circuit. In 1950, the scientists started applying their idea to
work and analyzed how human neurons might work.

17 18

Prepared by: Dr. Rabins Porwal 3

 Machine intelligence in Games: The first "AI" winter:
• 1952: Arthur Samuel, who was the pioneer of machine learning, • The duration of 1974 to 1980 was the tough time for AI and ML
created a program that helped an IBM computer to play a checkers researchers, and this duration was called as AI winter.
game. It performed better more it played. • In this duration, failure of machine translation occurred, and people
• 1959: In 1959, the term "Machine Learning" was first coined by had reduced their interest from AI, which led to reduced funding by
Arthur Samuel. the government to the researches.

19 20

Machine Learning from theory to reality Machine Learning at 21st century

• 2006: In the year 2006, computer scientist Geoffrey Hinton has given a new name
• 1959: In 1959, the first neural network was applied to a real-world to neural net research as "deep learning," and nowadays, it has become one of the
problem to remove echoes over phone lines using an adaptive filter. most trending technologies.
• 2012: In 2012, Google created a deep neural network which learned to recognize
• 1985: In 1985, Terry Sejnowski and Charles Rosenberg invented a the image of humans and cats in YouTube videos.
neural network NETtalk, which was able to teach itself how to • 2014: In 2014, the Chabot "Eugen Goostman" cleared the Turing Test. It was the
correctly pronounce 20,000 words in one week. first Chabot who convinced the 33% of human judges that it was not a machine.
• 1997: The IBM's Deep blue intelligent computer won the chess game • 2014: DeepFace was a deep neural network created by Facebook, and they claimed
against the chess expert Garry Kasparov, and it became the first that it could recognize a person with the same precision as a human can do.
computer which had beaten a human chess expert. • 2016: AlphaGo beat the world's number second player Lee sedol at Go game. In
2017 it beat the number one player of this game Ke Jie.
• 2017: In 2017, the Alphabet's Jigsaw team built an intelligent system that was able
to learn the online trolling. It used to read millions of comments of different
websites to learn to stop online trolling

21 22

Machine Learning at present: Prerequisites

• Now machine learning has got a great advancement in its research, • Before learning machine learning, you must have the basic knowledge
and it is present everywhere around us, such as self-driving cars, of followings so that you can easily understand the concepts of
Amazon Alexa, Catboats, recommender system, and many more. It machine learning:
includes Supervised, unsupervised, and reinforcement learning with • Fundamental knowledge of probability and linear algebra.
clustering, classification, decision tree, SVM algorithms, etc. • The ability to code in any computer language, especially in Python language.
• Modern machine learning models can be used for making various • Knowledge of Calculus, especially derivatives of single variable and
predictions, including weather prediction, disease prediction, stock multivariate functions.
market analysis, etc.

23 24

Prepared by: Dr. Rabins Porwal 4