lOMoARcPSD|32645978

My Internship Report - This is my project work.

Artificial Intelligence Lab (Dev Bhoomi Institute of Technology Dehradun)

Studocu is not sponsored or endorsed by any college or university

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)
 lOMoARcPSD|32645978

DEV BHOOMI
UTTARAKHAND
UNIVERSITY

EIGHT WEEKS SUMMER INTERNSHIP

REPORT
On
Machine Learning with Python (Basics)

Submitted By:
RISHABH TYAGI
Under the guidance of my mentors and my respected
faculties.
SoCSE
July 2022 – August 2022

Page 1 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

DECLARATION
I hereby declare that I have completed my eight weeks summer
training at YHills (one of the country’s leading online
certification training providers) from 1st July, 2022 to 31st
August, 2022 under the guidance of my tutor. I have declared
that I have worked with full dedication during these eight weeks
of training and my learning outcomes fulfil the requirements of
training for the award of degree of Bachelor of Science (B.Sc) in
IT, Dev Bhoomi Uttarakhand University.

__________________
(Signature of Student)
Name – Rishabh Tyagi
ID – 21BSIT0004

Date – 28/Sep/2022

Page 2 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

ACKNOWLEDGEMENT
The success and final outcome of learning Machine Learning
required a lot of guidance and assistance from many people and I
am extremely privileged to have got this all along the completion
of my course and few of the projects. All that I have done is only
due to such supervision and assistance and I would not forget to
thank them.
I respect and thank YHills, for providing me an opportunity to
do the course and project work and giving me all support and
guidance, which made me complete the course duly. I am
extremely thankful to the course advisor Miss. Lata Thiru
Ma’am.
I am thankful to and fortunate enough to get constant
encouragement, support and guidance from all Teaching staffs of
YHills which helped me in successfully completing my course
and project work.

__________________
(Signature of Student)
Name – Rishabh Tyagi
ID – 21BSIT0004

Date – 28/Sep/2022

Page 3 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

[ SUMMER TRAINING CERTIFICATES ]

Page 4 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

TABLE OF CONTENTS
1. Introduction………………………………………………………………………………07
1.1. A Taste of Machine Learning……………………………………………………….07
1.2. Relation to Data Mining………………………………………………………….….07
1.3. Relation to Optimization………………………………………………………….…07
1.4. Relation to Statistics…………………………………………………………............08
1.5. Future of Machine Learning………………………………………………………....08
2. Technology Learnt……………………………………………………………………….08
2.1. Introduction to Artificial Intelligence and Machine Learning……………………....08
2.1.1. Definition of Artificial Intelligence…………………………………………..08
2.1.2. Definition of Machine Learning…………………………………………...…09
2.1.3. Machine Learning Algorithms……………………………………………….10
2.1.4. Applications of Machine Learning………………………………………...…11
2.2. Techniques of Machine Learning…………………………………………………....12
2.2.1. Supervised Learning…………………………………………...……………..12
2.2.2. Unsupervised Learning……………………………………………...………..16
2.2.3. Semi- supervised Learning……………………………………………..…….18
2.2.4. Reinforcement Learning…………………………………………………..….19
2.2.5. Some Important Considerations in Machine Learning………………….........19
2.3. Data Preprocessing………………………………………………………….…….....20
2.3.1. Data Preparation…………………………………………………………..….20
2.3.2. Feature Engineering…………………………………………………….….…21
2.3.3. Feature Scaling…………………………………………………………….…22
2.3.4. Datasets………………………………………………………………….……24
2.3.5. Dimensionality Reduction with Principal Component Analysis………….….24
2.4. Math Refresher………………………………………………………………………25
2.4.1. Concept of Linear Algebra……………………………………………...……25
2.4.2. Eigenvalues, Eigenvectors, and Eigen decomposition…………………….....30
2.4.3. Introduction to Calculus…………………………………………………..….30
2.4.4. Probability and Statistics………………………………………………….….31
2.5. Supervised learning……………………………………………………………….…34
2.5.1. Regression……………………………………………………………………34
2.5.1.1. Linear Regression…………………………………………………….35
2.5.1.2. Multiple Linear Regression…………………………………………..35
2.5.1.3. Polynomial Regression……………………………………………….36
2.5.1.4. Decision Tree Regression…………………………………………….37
2.5.1.5. Random Forest Regression…………………………………………...37
2.5.2. Classification…………………………………………………………………38
2.5.2.1. Linear Models…………………………………………………….…..39
2.5.2.1.1. Logistic Regression…………………………………………..39
2.5.2.1.2. Support Vector machines…………………………………….39

Page 5 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.5.2.2. Nonlinear Models…………………………………………………….40

2.5.2.2.1. K-Nearest Neighbors (KNN)…………………………………40
2.5.2.2.2. Kernel Support Vector Machines (SVM)…………………….40
2.5.2.2.3. Naïve Bayes…………………………………………………..41
2.5.2.2.4. Decision Tree Classification……………………………….…41
2.5.2.2.5. Random Forest Classification…………………………….…..42
2.6. Unsupervised learning…………………………………………………………….…43
2.6.1. Clustering…………………………………………………………………….43
2.6.1.1. Clustering Algorithms………………………………………………..43
2.6.1.2. K-means Clustering…………………………………………………..44
2.7. Introduction to Deep Learning………………………………………………………45
2.7.1. Meaning and Importance of Deep Learning………………………………….45
2.7.2. Artificial Neural Networks…………………………………………………...46
2.7.3. TensorFlow…………………………………………………………………...47
3. Reason for choosing Machine Learning………………………………………………….47
4. Learning Outcome………………………………………………………………………..48
5. Gantt Chart……………………………………………………………………………….49
6. Bibliography……………………………………………………………………………...49
6.1. Content source……………………………………………………………………….49
6.2. Picture from………………………………………………………………………….49
6.3. Book referred………………………………………………………………………..49

Page 6 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

1. Introduction
1.1. A Taste of Machine Learning
 Arthur Samuel, an American pioneer in the field of computer gaming and artificial intelligence, coined
the term "Machine Learning" in 1959.
 Over the past two decades Machine Learning has become one of the mainstays of information
technology.
 With the ever-increasing amounts of data becoming available there is good reason to believe that smart
data analysis will become even more pervasive as a necessary ingredient for technological progress.
1.2. Relation to Data Mining

• Data mining uses many machine learning methods, but with different goals; on the other hand,
machine learning also employs data mining methods as "unsupervised learning" or as a
preprocessing step to improve learner accuracy.
1.3. Relation to Optimization

Machine learning also has intimate ties to optimization: many learning problems are formulated as
minimization of some loss function on a training set of examples.

Loss functions express the discrepancy between the predictions of the model being trained and the
actual problem instances.

Page 7 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

1.4. Relation to Statistics

Michael I. Jordan suggested the term dat a science as a placeholder to call the overall
field.

Leo Breiman distinguished two statistical modelling paradigms: data model and algorithmic model,
wherein "algorithmic model" means more or less the machine learning algorithms like Random
forest.
1.5. Future of Machine Learning
 Machine Learning can be a competitive advantage to any company be it a top MNC or a startup as
things that are currently being done manually will be done tomorrow by machines.
 Machine Learning revolution will stay with us for long and so will be the future of Machine Learning.
2. Technology Learnt
2.1. Introduction to AI & Machine Learning
2.1.1. Definition of Artificial Intelligence
 Data Economy
 World is witnessing real time flow of all types structured and unstructured data from social media,
communication, transportation, sensors, and devices.
 International Data Corporation (IDC) forecasts that 180 zettabytes of data will be generated by 2025.

 This explosion of data has given rise to a new economy known as the Data Economy.

Page 8 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Data is the new oil that is precious but useful only when cleaned and processed.
 There is a constant battle for ownership of data between enterprises to derive benefits from it.
 Define Artificial Intelligence
Artificial intelligence refers to the simulation of human intelligence in machines that are programmed to think
like humans and mimic their actions. The term may also be applied to any machine that exhibits traits associated
with a human mind such as learning and problemsolving.

2.1.2. Definition of Machine Learning

 Relationship between AI and ML

Machine Learning is an approach or subset of Artificial Intelligence that is based on the idea that machines can be
given access to data along with the ability to learn from it.

Page 9 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Define Machine Learning

Machine learning is an application of artificial intelligence (AI) that provides systems the ability to
automatically learn and improve from experience without being explicitly programmed. Machine learning
focuses on the development of computer programs that can access data and use it learn for themselves.

 Features of Machine Learning

 Machine Learning is computing-intensive and generally requires a large amount of training data.
 It involves repetitive training to improve the learning and decision making of algorithms.
 As more data gets added, Machine Learning training can be automated for learning new data
patterns and adapting its algorithm.
2.1.3. Machine Learning Algorithms
 Traditional Programming vs. Machine Learning Approach

 Traditional Approach
Traditional programming relies on hard-coded rules.

 Machine Learning Approach

Machine Learning relies on learning patterns based on sample data.

Page 10 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Machine Learning Techniques

 Machine Learning uses a number of theories and techniques from Data Science.

 Machine Learning can learn from labelled data (known as supervised learning) or unlabeled
data (known as unsupervised learning).

Page 11 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.1.4. Applications of Machine Learning

 Image Processing
 Optical Character Recognition (OCR)
 Self-driving cars
 Image tagging and recognition
 Robotics
 Industrial robotics
 Human simulation
 Data Mining
 Association rules
 Anomaly detection
 Grouping and Predictions
 Video games
 Pokémon
 PUBG
 Text Analysis
 Spam Filtering
 Information Extraction
 Sentiment Analysis
 Healthcare
 Emergency Room & Surgery
 Research
 Medical Imaging & Diagnostics
2.2. Techniques of Machine Learning
2.2.1. Supervised Learning
❖ Define Supervised Learning

Supervised learning is the machine learning task of learning a function that maps an input to an output based on
example input-output pairs. It infers a function from labeled training data consisting of a set of training
examples.

Page 12 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

In supervised learning, each example is a pair consisting of an input object (typically a vector) and a desired
output value (also called the supervisory signal).

❖ Supervised Learning Flow

 Data Preparation Clean data
Label data (x, y)

Feature Engineering

Reserve 80% of data for Training (Train_X) and 20% for Evaluation (Train_E)

 Training Step
Design algorithmic logic

Train the model with Train X

Derive the relationship between x and y, that is, y = f(x)

 Evaluation or Test Step

Evaluate or test with Train E

If accuracy score is high, you have the final learned algorithm y = f(x) If accuracy
score is low, go back to training step

 Production Deployment
Use the learned algorithm y = f(x) to predict production data.

The algorithm can be improved by more training data, capacity, or algo redesign.

❖ Testing the Algorithms

 Once the algorithm is trained, test it with test data (a set of data instances that do not appear in
the training set).
 A well-trained algorithm can predict well for new test data.

Page 13 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 If the learning is poor, we have an underfitted situation. The algorithm will not work well on
test data. Retraining may be needed to find a better fit.

 If learning on training data is too intensive, it may lead to overfitting–a situation where the
algorithm is not able to handle new testing data that it has not seen before. The technique to
keep data generic is called regularization.

Page 14 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Examples of Supervised Learning

 Voice Assistants
 Gmail Filters
 Weather Apps
 Types of Supervised Learning

 Classification
 Answers <What class?=

Page 15 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Applied when the output has finite and discreet values Example: Social media
sentiment analysis has three potential outcomes, positive, negative, or neutral
✓ Regression

 Answers <How much?=

 Applied when the output is a continuous number

 A simple regression algorithm: y = wx + b. Example: relationship
between environmental temperature (y) and humidity levels (x)
2.2.2. Unsupervised Learning

 Define Unsupervised Learning

Unsupervised learning is the training of machine using information that is neither classified nor labeled and
allowing the algorithm to act on that information without guidance.

Page 16 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

Here the task of machine is to group unsorted information according to similarities, patterns and differences
without any prior training of data.

 Types of Unsupervised Learning

 Clustering

The most common unsupervised learning method is cluster analysis. It is used to find data clusters so that each
cluster has the most closely matched data.

Page 17 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Visualization Algorithms

Visualization algorithms are unsupervised learning algorithms that accept unlabeled data and display this data
in an intuitive 2D or 3D format. The data is separated into somewhat clear clusters to aid understanding.

 Anomaly Detection
This algorithm detects anomalies in data without any prior training.

2.2.3. Semi- supervised Learning

 Define Semi-supervised Learning
Semi-supervised learning is a class of machine learning tasks and techniques that also make use of unlabeled
data for training – typically a small amount of labeled data with a large amount of unlabeled data.

Semi-supervised learning falls between unsupervised learning (without any labeled training data) and
supervised learning (with completely labeled training data).

 Example of Semi-supervised Learning

 Google Photos automatically detects the same person in multiple photos from a vacation trip
(clustering –unsupervised).
 One has to just name the person once (supervised), and the name tag gets attached to that person
in all the photos.

Page 18 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.2.4. Reinforcement Learning

 Define Reinforcement Learning
Reinforcement Learning is a type of Machine Learning that allows the learning system to observe the
environment and learn the ideal behavior based on trying to maximize some notion of cumulative reward.

It differs from supervised learning in that labelled input/output pairs need not be presented, and sub-optimal
actions need not be explicitly corrected. Instead the focus is finding a balance between exploration (of
uncharted territory) and exploitation (of current knowledge)

 Features of Reinforcement Learning

• The learning system (agent) observes the environment, selects and takes certain actions, and
gets rewards in return (or penalties in certain cases).
• The agent learns the strategy or policy (choice of actions) that maximizes its rewards over time.
 Example of Reinforcement Learning
• In a manufacturing unit, a robot uses deep reinforcement learning to identify a device from one
box and put it in a container.
• The robot learns this by means of a rewards-based learning system, which incentivizes it for
the right action.
2.2.5. Some Important Considerations in Machine Learning
 Bias & Variance Tradeoff
 Bias refers to error in the machine learning model due to wrong assumptions. A high-bias model
will underfit the training data.
 Variance refers to problems caused due to overfitting. This is a result of oversensitivity of the
model to small variations in the training data. A model with many degrees of freedom (such as
a high-degree polynomial model) is likely to have high variance and thus overfit the training
data.
 Bias & Variance Dependencies
 Increasing a model’s complexity will reduce its bias and increase its variance.

Page 19 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Conversely, reducing a model’s complexity will increase its bias and reduce its variance. This
is why it is called a tradeoff.
 What is Representational Learning
In Machine Learning, Representation refers to the way the data is presented. This often make a huge difference
in understanding.

2.3. Data Preprocessing

2.3.1. Data Preparation
 Data Preparation Process
 Machine Learning depends largely on test data.
 Data preparation involves data selection, filtering, transformation, etc.

Data preparation is a crucial step to make it suitable for ML.

A large amount of data is generally required for the most common forms of ML.
 Types of Data
 Labelled Data or Training Data

Page 20 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Unlabeled Data
 Test Data
 Validation Data
2.3.2. Feature Engineering
 Define Feature Engineering
The transformation stage in the data preparation process includes an important step known as Feature
Engineering.

Feature Engineering refers to selecting and extracting right features from the data that are relevant to the task
and model in consideration.

 Aspects of Feature Engineering

 Feature Selection
Most useful and relevant features are selected from the available data

 Feature Addition
New features are created by gathering new data

 Feature Extraction

Existing features are combined to develop more useful ones

 Feature Filtering

Page 21 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

Filter out irrelevant features to make the modelling step easy

2.3.3. Feature Scaling

 Define Feature Scaling
 Feature scaling is an important step in the data transformation stage of data preparation process.
 Feature Scaling is a method used in Machine Learning for standardization of independent
variables of data features.
 Techniques of Feature Scaling

 Standardization
 Standardization is a popular feature scaling method, which gives data the property of
a standard normal distribution (also known as Gaussian distribution).
 All features are standardized on the normal distribution (a mathematical model).
 The mean of each feature is centered at zero, and the feature column has a standard
deviation of one.

Page 22 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

✓ Normalization

 In most cases, normalization refers to rescaling of data features between 0 and 1,

which is a special case of Min-Max scaling.

 In the given equation, subtract the min value for each feature from each feature
instance and divide by the spread between max and min.
 In effect, it measures the relative percentage of distance of each instance from the min
value for that feature.

Page 23 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.3.4. Datasets
 Machine Learning problems often need training or testing datasets.
 A dataset is a large repository of structured data.
 In many cases, it has input and output labels that assist in Supervised Learning.
2.3.5. Dimensionality Reduction with Principal Component Analysis
 Define Dimensionality Reduction
 Dimensionality reduction involves transformation of data to new dimensions in a way that
facilitates discarding of some dimensions without losing any key information.

 Define Principal Component Analysis (PCA)

 Principal component analysis (PCA) is a technique for dimensionality reduction that helps in
arriving at better visualization models.

Page 24 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Applications of PCA
 Noise reduction
 Compression
 Preprocess
2.4. Math Refresher
2.4.1. Concept of Linear Algebra
 Linear Equation
Linear algebra is a branch of mathematics that deals with the study of vectors and linear
functions and equations.

A linear equation does not involve any products, inverses, or roots of variables. All variables
occur only to the first power and not as arguments for trigonometric, logarithmic, or
exponential functions.

 System of Linear Equations

A system of linear equations is a finite collection of linear equations.

Page 25 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

A linear system that has a solution is called consistent, and the one with no solution is termed
inconsistent.

 Matrix

An m × n matrix: the m rows are horizontal and the n columns are vertical. Each element of a matrix is often
denoted by a variable with two subscripts. For example, a2,1 represents the element at the second row and first
column of the matrix.

 Addition

Two matrices can be added only if they have the same number of rows and columns. Also, during addition, A +
B=B+A

 Subtraction
Two matrices can be subtracted only if they have the same number of rows and columns. Also, during
subtraction, A -B not equal to B -A

 Multiplication

Page 26 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

The matrix product AB is defined only when the number of columns in A is equal to the number of rows in B.
BA is defined only when the number of columns in B is equal to the number of rows in A. AB is not always
equal to BA.

 Transpose

 Inverse

❖ Special Types of Matrix

 Diagonal Matrix

Page 27 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Symmetric Matrix

 Identity Matrix

❖ Vector

Page 28 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

A vector (v) is an object with both magnitude (length) and direction.

It starts from origin (0,0), and its length is denoted by ||v||.

 Addition
The operation of adding two or more vectors together into a vector sum is referred to
as vector addition.

 Subtraction
Vector subtraction is the process of subtracting two or more vectors to get a vector
difference.

 Multiplication
Vector multiplication refers to a technique for the multiplication of two (or more)
vectors with themselves.

Page 29 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.4.2. Eigenvalues, Eigenvectors, and Eigen decomposition

 Eigenvalue & Eigenvector
 An eigenvector of a square matrix A is a non-zero vector such that multiplication by A alters
only the scale of v.

 Eigen decomposition
 Integers can be broken into their prime factors to understand them, example: 12 = 2 x 2 x 3.
From this, useful properties can be derived, for example, the number is not divisible by 5 and
is divisible by 2 and 3.
 Similarly, matrices can be decomposed. This will help you discover information about the
matrix.

2.4.3. Introduction to Calculus

Calculus is the study of change. It provides a framework for modelling systems in which there is change and
ways to make predictions of such models.

 Differential Calculus
 Differential calculus is a part of calculus that deals with the study of the rates at which quantities
change.
 Let x and y be two real numbers such that y is a function of x, that is, y = f(x).
 If f(x) is the equation of a straight line (linear equation), then the equation is represented as y =
mx + b.
 Where m is the slope determined by the following equation:

Page 30 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Integral Calculus
 Integral Calculus assigns numbers to functions to describe displacement, area, volume, and
other concepts that arise by combining infinitesimal data.
 Given a function f of a real variable x and an interval [a, b] of the real line, the definite integral
is defined informally as the signed area of the region in the xyplane that is bounded by the graph
of f, the x -axis, and the vertical lines x=a and x=b.

2.4.4. Probability and Statistics

 Probability Theory
 Probability is the measure of the likelihood of an event’s occurrence. ➢ Example: The
chances of getting heads on a coin toss is ½ or 50%

 Probability of any specific event is between 0 and 1 (inclusive). The sum of total probabilities
of an event cannot exceed 1, that is, 0 <= p(x) <= 1. This implies that. p(x)dx =1 (integral of p
for a distribution over x)
 Conditional Probability
 Conditional Probability is a measure of the probability of an event occurring given that another
event has occurred.

Page 31 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Chain Rule of Probability

 Joint probability distribution over many random variables may be decomposed into conditional
distributions over only one variable. ➢ It can be represented as:

 Standard Deviance
 Standard deviation is a quantity that expresses the value by which the members of a group differ
from the mean value for the group.

 Standard deviation is used more often than variance because the unit in which it is measured is
the same as that of mean, a measure of central tendency.
 Variance
 Variance refers to the spread of the data set, for example, how far the numbers are in relation
to the mean.
 Variance is particularly useful when calculating the probability of future events or performance.

 Notice that variance is just the square of standard deviation.

 Covariance
 Covariance is the measure of how two random variables change together. It is used to calculate
the correlation between variables.

Page 32 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Logistic Sigmoid
 The Logistic Sigmoid is a useful function that follows the S curve. It saturates when input is
very large or very small.

 Gaussian Distribution
 The distribution where the data tends to be around a central value with lack of bias or minimal
bias toward the left or right is called Gaussian distribution, also known as normal distribution.

Page 33 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.5. Supervised learning

2.5.1. Regression
 In statistical modeling, regression analysis is a set of statistical processes for estimating the relationships
among variables.
 It includes many techniques for modeling and analyzing several variables, when the focus is on the
relationship between a dependent variable and one or more independent variables (or 'predictors').
 More specifically, regression analysis helps one understand how the typical value of the dependent
variable (or 'criterion variable') changes when any one of the independent variables is varied, while the
other independent variables are held fixed.

Page 34 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.5.1.1. Linear Regression

 Linear regression is a linear approach for modeling the relationship between a scalar
dependent variable y and an independent variable x.

 where x, y, w are vectors of real numbers and w is a vector of weight parameters.

 The equation is also written as:

y = wx + b
 where b is the bias or the value of output for zero input
2.5.1.2. Multiple Linear Regression
It is a statistical technique used to predict the outcome of a response variable through
several explanatory variables and model the relationships between them.

Page 35 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

It represents line fitment between multiple inputs and one output, typically:

2.5.1.3. Polynomial Regression

• Polynomial regression is applied when data is not formed in a straight line.

• It is used to fit a linear model to non-linear data by creating new features from powers of non-linear
features.

Page 36 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.5.1.4. Decision Tree Regression

o A decision tree is a graphical representation of all the possible solutions to a decision based
on a few conditions. o Decision Trees are non-parametric models, which means that the
number of parameters is not determined prior to training. Such models will normally
overfit data.
o In contrast, a parametric model (such as a linear model) has a predetermined number of
parameters, thereby reducing its degrees of freedom. This in turn prevents overfitting.

o max_depth –limit the maximum depth of the tree

o min_samples_split –the minimum number of samples a node must have before it can be
split
o min_samples_leaf –the minimum number of samples a leaf node must have o
min_weight_fraction_leaf –same as min_samples_leaf but expressed as a fraction of total
instances
o max_leaf_nodes –maximum number of leaf nodes
o max_features –maximum number of features that are evaluated for splitting at each node
2.5.1.5. Random Forest Regression
 Ensemble Learning uses the same algorithm multiple times or a group of different
algorithms together to improve the prediction of a model.

Page 37 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Random Forests use an ensemble of decision trees to perform regression tasks.

2.5.2. Classification
It specifies the class to which data elements belong to.
It predicts a class for an input variable.
It is best used when the output has f inite and discreet values.

There are 2 types of classification, binomial and multi - class .

Page 38 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.5.2.1. Linear Models

2.5.2.1.1. Logistic Regression
 This method is widely used for binary classification problems. It can also be extended
to multi-class classification problems.
 A binary dependent variable can have only two values, like 0 or 1, win or lose, pass
or fail, healthy or sick, etc.

 The probability in the logistic regression is often represented by the Sigmoid function
(also called the logistic function or the S-curve)

2.5.2.1.2. Support Vector machines

 SVMs are very versatile and are also capable of performing linear or nonlinear
classification, regression, and outlier detection.
 They involve detecting hyperplanes which segregate data into classes.

Page 39 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 The optimization objective is to find <maximum margin hyperplane= that is farthest from
the closest points in the two classes (these points are called support vectors).
2.5.2.2. Nonlinear Models
2.5.2.2.1. K-Nearest Neighbors (KNN)
K-nearest Neighbors algorithm is used to assign a data point to clusters based on
similarity measurement.

A new input
point is classified in the category such that it has the greatest number of neighbors
from that category.
2.5.2.2.2. Kernel Support Vector Machines (SVM)
Kernel SVMs are used for classification of nonlinear data.

In the chart, nonlinear data is projected into a higher dimensional space via a mapping
function where it becomes linearly separable.

Page 40 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

A
reverse

projection of the higher dimension back to original feature space takes it back to
nonlinear shape.
2.5.2.2.3. Naïve Bayes
 According to Bayes model, the conditional probability P(Y|X) can be calculated as:

 This means you have to estimate a very large number of P(X|Y) probabilities for a
relatively small vector space X.

2.5.2.2.4. Decision Tree Classification

 The advantage of decision trees is that they require very little data preparation.
 They do not require feature scaling or centering at all.
 They are also the fundamental components of Random Forests, one of the most
powerful ML algorithms.

Page 41 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 Start at the tree root and split the data on the feature using the decision algorithm,
resulting in the largest information gain (IG).
2.5.2.2.5. Random Forest Classification
 Random decision forests correct for decision trees' habit of overfitting to their training set.

 Random forests or random decision forests are an ensemble learning method for
classification, regression and other tasks that operates by constructing a multitude of
decision trees at training time and outputting the class that is the mode of the classes
(classification) or mean prediction (regression) of the individual trees.

Page 42 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.6. Unsupervised learning

2.6.1. Clustering
2.6.1.1. Clustering Algorithms

 Clustering means
✓ Clustering is a Machine Learning technique that involves the grouping of data points.

 Prototype Based Clustering

 Prototype-based clustering assumes that most data is located near prototypes;
example: centroids (average) or medoid (most frequently occurring point)
 K-means, a Prototype-based method, is the most popular method for clustering that
involves:
• Training data that gets assigned to matching cluster based on similarity
• Iterative process to get data points in the best clusters possible

Page 43 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

2.6.1.2. K-means Clustering

 K-means Clustering Algorithm

Step 1: randomly pick k centroids

Step 2: assign each point to the nearest centroid

Step 3: move each centroid to the center of the respective cluster

Step 4: calculate the distance of the centroids from each point again

Step 5: move points across clusters and re-calculate the distance from the centroid

Step 6: keep moving the points across clusters until the Euclidean distance is
minimized

 Elbow Method
 One could plot the Distortion against the number of clusters K. Intuitively, if K increases,
distortion should decrease. This is because the samples will be close to their assigned centroids.
This plot is called the Elbow method.

Page 44 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 It indicates the optimum number of clusters at the position of the elbow, the point where
distortion begins to increase most rapidly.
 Euclidian Distance
✓ K-means is based on finding points close to cluster centroids. The distance between
two points x and y can be measured by the squared Euclidean distance between them
in an m-dimensional space.

 Examples of K-means Clustering

 Grouping articles (example: Google news)
 Grouping customers who share similar interests
 Classifying high risk and low risk patients from a patient pool
2.7. Introduction to Deep Learning
2.7.1. Meaning and Importance of Deep Learning
 Define Deep Learning
Deep Learning is a specialized form of Machine Learning that uses supervised, unsupervised, or semi-
supervised learning to learn data representations.

It is similar to the structure and function of the human nervous system.

 Why Deep Learning

The vast availability of Big Data enables machines to be trained.

Experts have discovered multi-layered learning networks that can be leveraged for deep learning
as they learn in layers.

Scientists have figured out that high-performing graphics processing units (GPU) can be used for
deep learning.

Page 45 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 ML Vs Deep Learning

2.7.2. Artificial Neural Networks

 Deep learning relies on multiple layers of training.
 Artificial Neural Network is a computing system made up of a number of simple, highly
interconnected processing elements which process information by their dynamic state response
to external inputs.

 It is an interconnected group of nodes akin to the vast network of layers of neurons in a brain.
2.7.3. TensorFlow
 TensorFlow is the open source Deep Learning library provided by Google.

Page 46 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

 It allows development of a variety of neural network applications such as computer vision, speech
processing, or text recognition.
 It uses data flow graphs for numerical computations.
3. Reason for choosing Machine Learning
➢ Learning machine learning brings in better career opportunities
 Machine learning is the shining star of the moment.
 Every industry looking to apply AI in their domain, studying machine learning opens world of
opportunities to develop cutting edge machine learning applications in various verticals – such
as cyber security, image recognition, medicine, or face recognition.
 Several machine learning companies on the verge of hiring skilled ML engineers, it is becoming
the brain behind business intelligence.
➢ Machine Learning Jobs on the rise
 The major hiring is happening in all top tech companies in search of those special kind of people
(machine learning engineers) who can build a hammer (machine learning algorithms).
 The job market for machine learning engineers is not just hot but it’s sizzling.
 Machine Learning Jobs on Indeed.com - 2,500+(India) & 12,000+(US)

Page 47 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

4. Learning Outcome
 Have a good understanding of the fundamental issues and challenges of machine learning: data, model
selection, model complexity, etc.
 Have an understanding of the strengths and weaknesses of many popular machine learning approaches.
 Appreciate the underlying mathematical relationships within and across Machine Learning algorithms
and the paradigms of supervised and un-supervised learning.
 Be able to design and implement various machine learning algorithms in a range of real-world
applications.
 Ability to integrate machine learning libraries and mathematical and statistical tools with modern
technologies
 Ability to understand and apply scaling up machine learning techniques and associated computing
techniques and technologies.

Page 48 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)

 lOMoARcPSD|32645978

5. Gantt Chart

6. Bibliography
6.1. All Content used in this report is from
 https://www.simplilearn.com/
 https://www.wikipedia.org/
 https://towardsdatascience.com/
 https://www.expertsystem.com/
 https://www.coursera.org/
 https://www.edureka.co/
 https://subhadipml.tech/
 https://www.forbes.com/
 https://medium.com/
 https://www.google.com/
6.2. All Pictures are from
 https://www.simplilearn.com/
 https://www.google.com/
 https://www.wikipedia.org/
 https://www.youtube.com/
 https://www.edureka.co/
6.3. Book I referred are

 Hands-on Machine Learning with Scikit-learn & Tensorflow By Aurelien Geron

 Python Machine Learning by Sebastian Raschka

Page 49 of 49

Downloaded by Arun kumar Dhulipalla (21jr1a4362@gmail.com)