Acropolis Institute of Technology and Research

Neha Sehta Introduction to Machine Learning 1 / 23

 Introduction to Machine Learning

Neha Sehta

Neha Sehta Introduction to Machine Learning 2 / 23

What is Learning?

The activity or process of gaining knowledge or skill by

studying, practicing, being taught, or experiencing
something.

Neha Sehta Introduction to Machine Learning 3 / 23

What is Machine Learning?

Machine learning is one way to bring AI. It was defined in the 1950s by AI
pioneer Arthur Samuel as
“the field of study that gives computers the ability to learn without
explicitly being programmed.”

Neha Sehta Introduction to Machine Learning 4 / 23

What is Artificial Intelligence?

Making machines portray intelligent behavior.

Making machines mimic human behavior.
▶ Recognition(Face, Speech, Context, Emotions)
▶ Reasoning
▶ Interpreting
▶ Game Playing
▶ Planning
▶ Searching
▶ Inference
▶ Learning
▶ all that where humans are still better than machines.

Neha Sehta Introduction to Machine Learning 5 / 23

Learning vs ML

Neha Sehta Introduction to Machine Learning 6 / 23

AI Objectives

Neha Sehta Introduction to Machine Learning 7 / 23

ML Vs Statistics

It’s similar to statistics...

▶ Both fields try to uncover patterns in data
▶ Both fields draw heavily on calculus, probability, and linear algebra, and
share many of the same core algorithms
But it’s not statistics!
▶ Stats is more concerned with helping scientists and policymakers draw
good conclusions; ML is more concerned with building autonomous
agents
▶ Stats puts more emphasis on interpretability and mathematical rigor;
ML puts more emphasis on predictive performance, scalability, and
autonomy

Neha Sehta Introduction to Machine Learning 8 / 23

Definitions

Definition 1
A computer program is said to learn from experience E with respect to
some task T and some performance measure P, If its performance on T, as
measured by P improves with experience E.- Tom Mitchell

Definition 2
Machine learning is a branch of artificial intelligence (AI) and computer
science which focuses on the use of data and algorithms to imitate the way
that humans learn, gradually improving its accuracy.

Definition 3
Machine learning is an application of AI that enables systems to learn and
improve from experience without being explicitly programmed. Machine
learning focuses on developing computer programs that can access data
and use it to learn for themselves.
Neha Sehta Introduction to Machine Learning 9 / 23
AI vs ML vs DL

Neha Sehta Introduction to Machine Learning 10 / 23

Traditional programming Vs Machine Learning

Neha Sehta Introduction to Machine Learning 11 / 23

Data Science Vs Machine Learning

Data science is a field that studies data and how to extract meaning
from it, whereas machine learning is a field devoted to understanding
and building methods that utilize data to improve performance or
inform predictions. Machine learning is a branch of artificial
intelligence.
Machine learning automates the process of data analysis and goes
further to make predictions based on collecting and analyzing large
amounts of data on certain populations. Models and algorithms are
built to make this happen.
Data Mining: Data mining is the process of sorting through large
data sets to identify patterns and relationships that can help solve
business problems through data analysis.

Neha Sehta Introduction to Machine Learning 12 / 23

Scope of Machine Learning- Top Use Cases

1. Health Care
ML programs reading patient history, records, reports etc. to devise
personalised treatment plans. IBM Watson Oncology is an important
project in this space.
Wearable technology for disease prevention and elder healthcare
monitoring is also making great strides.
ML is used to analyze massive healthcare data sets to accelerate
discovery of treatments and cures, improve patient outcomes, and
automate routine processes like blood test etc to prevent human error.
Machine learning can help with the diagnosis of diseases. Oncology
and pathology use machine learning to recognise cancerous tissue,
check body fluids etc
Robots performing complicated surgeries precisely.

Neha Sehta Introduction to Machine Learning 13 / 23

Scope of Machine Learning- Top Use Cases

2. Fraud Prevention
ML is being used in the financial and banking sector to autonomously
analyze large numbers of transactions to uncover fraudulent activity
in real time.
Classification and regression methods are being used to identify and
filter out phishing emails and spam mails.
Machine learning and computer vision algorithms are checking for
identity matching across key databases in real-time to prevent identity
theft.
ML techniques are also used to identify fake documents to prevent
forgery.

Neha Sehta Introduction to Machine Learning 14 / 23

Scope of Machine Learning- Top Use Cases
3. Product Recommendations
AI researchers and developers are using ML algorithms to develop AI
recommendation engines that offer relevant product suggestions
based on buyers’ past choices, as well as historical, geographic and
demographic data.
Products in a departmental stores are arranged such that they match
their association with one another to improve sales

4. Search Engine(SE) Result Refining

Google and other SEs use ML to improve the search results. Every
time you execute a search, the algorithms at the backend keep a
watch how u respond to the results. If you open the top results and
stay there for long, the SE assumes that the results were in
accordance to the query, But if you reach the second or third page of
results or do not open any of the results, the SE estimates that the
results served did not match requirement.
Neha Sehta Introduction to Machine Learning 15 / 23
Scope of Machine Learning- Top Use Cases

5. Social Media Services

Suggest Friends
personalizing your news feed
better ads targeting
Face Recognition - You upload a picture of you with a friend,
Facebook instantly recognizes that friend.

6. Predictions while Commuting

Traffic Predictions - ML helps to estimate the regions where
congestion can be found on the basis of daily experiences.
Online Transportation Networks: When booking a cab/food delivery,
the app based on ML estimates the price of the ride.

Neha Sehta Introduction to Machine Learning 16 / 23

Scope of Machine Learning- Top Use Cases

7. Virtual Personal Assistance

Smart car/phone/TV
Siri/Alexa/Google

8. Videos Surveillance
Imagine a single person monitoring multiple video cameras! Certainly,
a difficult job to do and boring as well. This is why the idea of
training computers to do this job makes sense.
The video surveillance system based on ML makes it possible to
detect crime before they happen. They track unusual behaviour like
standing motionless for a long time, stumbling, or napping on benches
etc. The system thus give an alert to human attendants, which can
ultimately help to avoid mishaps. And when such activities are
reported and counted to be true, they help to improve the surveillance
services.
Neha Sehta Introduction to Machine Learning 17 / 23
Limitations

1 ML is not based on knowledge

Machines are driven by data, not human knowledge. As a result,
“intelligence” is dictated by the volume of data you have to train it
with.
2 ML models are difficult to train
Eighty-one percent of data scientists admit that training with data is
more difficult than expected. It takes time and resources to train
machines. Massive data sets are needed to create data models, and
the process involves manually pre-tagging and categorizing data sets.
This resource drain can create latency and bottlenecks in advancing
ML initiatives.

Neha Sehta Introduction to Machine Learning 18 / 23

Limitations

3 Machine learning is prone to data issues

Machine learning is prone to data issues. Ninety-six percent of
companies have experienced training-related problems with data
quality, data labeling and building model confidence. Those
training-related problems are a key reason why seventy-eight percent
of ML projects stall prior to deployment. This has created an
extraordinarily high threshold for ML success.
4 Machine learning is often biased
Machine learning systems are known for operating in a black box,
meaning you have no visibility into how the machine learns and makes
decisions. Thus, if you identify an instance of bias, there is no way to
identify what caused it. Your only recourse is to retrain the algorithm
with additional data, but that is no guarantee to resolve the issue.

Neha Sehta Introduction to Machine Learning 19 / 23

When do we use ML?

1 ML is used when :
1 Human expertise does not exist (navigating on Mars)
2 Humans can’t explain their expertise (speech recognition)
3 Models must be customized (personalized medicine)
4 Models are based on huge amounts of data (genomics)

Neha Sehta Introduction to Machine Learning 20 / 23

AI but not ML

Expert systems
Evolutionary algorithms
Fuzzy logic
Planning and scheduling
Robotics: Robots are machines that can perform tasks autonomously
or semi-autonomously, by following pre-programmed instructions or by
responding to sensor inputs.
Computer Vision
Natural Language Processing

Neha Sehta Introduction to Machine Learning 21 / 23

ML but not AI

Neha Sehta Introduction to Machine Learning 22 / 23

First Exercise

Neha Sehta Introduction to Machine Learning 23 / 23