INTRODUCTION TO MACHINE LEARNING

UNIT–I:
Introduction:
Machine learning is a subset of artificial intelligence (AI) that involves the use of algorithms and statistical models to
enable computer systems to learn from data without being explicitly programmed. The process of machine learning
involves feeding large amounts of data into a machine learning model, which then analyzes the data and finds patterns
and relationships that can be used to make predictions or decisions. The model is then refined through a process called
training, in which it is repeatedly exposed to new data until it can accurately predict outcomes on its own. There are
several types of machine learning, including supervised learning, unsupervised learning, semi-supervised learning, and
reinforcement learning. Supervised learning involves training a model using labeled data where the desired output is
already known, such as predicting whether an email is spam or not based on previous labeled examples. Unsupervised
learning involves training a model using unlabeled data where the desired output is unknown, such as clustering similar
documents together based on their content. When training a model, semi-supervised learning uses both labeled and
unlabeled data, while reinforcement learning teaches a model to make decisions based on rewards or punishments,
which is similar to how people learn by making mistakes. Machine learning can be used for many things, like
recognizing images and voices, processing natural language, making recommendations, and finding fraud. It is also
used in fields such as healthcare, finance, and transportation to make predictions and improve decision-making
processes.

What is Machine Learning

Machine learning is a subfield of artificial intelligence (AI) that focuses on developing algorithms and models that
enable computer systems to learn from and improve on data without being explicitly programmed. The goal of machine
learning is to create systems that can automatically identify patterns, relationships, and insights in data and use this
information to make predictions or take actions. At its core, machine learning involves developing mathematical
models that can learn from data and make predictions or decisions based on what they have learned. These models are
trained on large datasets that contain examples of the problems they are designed to solve, such as classifying images,
detecting anomalies in data, or predicting the likelihood of a customer buying a product. Models for machine learning
can be put into a number of groups, such as supervised learning, unsupervised learning, semi-supervised learning, and
reinforcement learning.In supervised learning, models are trained on data that has been labeled, and the correct answers
are given to the model while it is being trained.Unsupervised learning is when models are trained on data that isn't
labeled, so the model has to figure out patterns and relationships on its own.Semi-supervised learning combines labeled
and unlabeled data to train models. Reinforcement learning involves training models to make decisions based on
rewards or punishments received as feedback. Machine learning has a lot of real-world uses, like recognizing images
and voices, processing natural language, finding fraud, making recommendations, and a lot more. It is used across
many different industries, including healthcare, finance, e-commerce, and more.

Use Machine Learning

Machine learning can be used in a wide variety of applications across many different industries. Here are a few
examples:
a) Image and speech recognition: machine learning algorithms can be trained to recognize objects in images or to
transcribe speech in real-time. This technology is used in applications like self-driving cars, virtual assistants,
and facial recognition systems.
b) Natural language processing: Machine learning can be used to understand and analyze human language. This
lets things like chatbots, voice assistants, and language translation tools work.
c) Fraud detection: Machine learning can be used to find patterns of fraud in financial transactions or other
activities. This lets fraud be found early and stopped.
d) Recommendation systems: Machine learning algorithms can be used to analyze user behavior and preferences
and provide personalized recommendations for products, services, or content.
e) Healthcare: In healthcare, machine learning is used to look at medical images, predict how a patient will do,
and help with diagnosis and treatment planning.
 f) Financial forecasting: machine learning algorithms can be used to predict stock prices, forecast market trends,
and analyze financial data.
g) Predictive maintenance: Machine learning algorithms can be used to monitor and analyze data from equipment
sensors to predict when maintenance is needed. This cuts down on downtime and maintenance costs.
These are just a few examples of how machine learning can be applied in different industries and applications. As the
technology keeps getting better and better, the number of things it can be used for will only grow.
Types of Machine Learning Systems
There are mainly three types of machine learning systems:
1) Supervised Learning: Supervised learning involves training a machine learning model on a labeled dataset
where the desired output is already known. The model is trained to make predictions based on the input features,
which can be used to classify or predict new data. Examples of supervised learning include image classification,
text classification, and regression analysis.
2) Unsupervised Learning: Unsupervised learning involves training a machine learning model on an unlabeled
dataset where the desired output is unknown. The model is tasked with finding patterns and relationships within
the data, such as grouping similar data points or identifying outliers. Examples of unsupervised learning include
clustering and anomaly detection.
3) Reinforcement Learning: Reinforcement learning involves training a machine learning model to make
decisions based on a reward or punishment system. The model learns through trial and error, receiving feedback
in the form of a reward when it makes a correct decision and a punishment when it makes an incorrect one.
Examples of reinforcement learning include game playing, robotics, and autonomous driving.
In machine learning, there are also subfields like semi-supervised learning, which uses both labeled and unlabeled data,
and deep learning, which uses neural networks to train models on large datasets. Each type of machine learning has its
own strengths and weaknesses, and the choice of which type to use depends on the specific problem being solved and
the availability of data.
Type of machine learning in which machine are trained using well labeled training data and machine predict the
output. Labeled data means some input data is already tagged with the correct output.

Types of Supervised learning

Classification

✔ Classification is a supervised learning

✔ Classification is a categorical variable
✔ Help you divide your data into different classes and the algorithm which implements the classification on
a dataset is known as a classifier.

✔ There are two types of classifications

 1) Binary classification: if the classification problem has only two possible classes is called binary
classification(T/F,Y/N,0,1)
2) Multi class classification: if the classification program has more than two classes is called multi class
classification(Movies, Music)

Types of Classification Algorithms

✔ Knn
✔ Naïve bayes
✔ Decision tree
✔ Logistic regression
✔ Support vector machine
Regression

✔ Regression algorithm is used if there is a relation between dependent and independent variable or input
and output variable is called regression.

✔ Regression it is used for the prediction of continuous variable such as a weather, forecasting, market
trends etc.

Types of Regression Algorithms

✔ Linear regression
✔ Logistic Regression
✔ Polynomial Regression
Unsupervised Learning

Unsupervised learning is a type of algorithm that learns patterns from untagged data. It mainly deal with
the unlabelled data Unsupervised learning algorithm allows users to perform more complex processing
task compared to supervised learning.

Clustering

Clustering is a unsupervised learning. There is not any label for each instance of data. Clustering is
alternatively called as grouping Clustering is the task of grouping a set of objects in such a way that objects
in the same group are more similar to each other than to those in other group.

Types of clustering algorithms

✔ Exclusive cluster
✔ Overlap cluster
✔ Hierarchical
 Reinforcement Learning

Reinforcement learning is an important type of machine learning where an agent

learns how to behave in an environment by performing actions and seeing the results.
Reinforcement learning is learning from mistakes at the beginning stage.
Reinforcement learning is a relationship between supervised and unsupervised
learning

Batch and Online Learning

Batch learning and online learning are two different approaches to machine learning that are used
to train models on data. Batch learning, also known as offline learning, involves training a model
on a fixed dataset or batch of data all at once. In batch learning, the entire dataset is processed, and
the model is trained on it to optimize the model parameters. This process requires a large amount
of computing power and storage, but the trained model can be used to make predictions on new
data quickly.
Online learning, which is also called incremental learning or streaming learning, involves changing
the model's parameters as new data comes in.Online learning works well when the information is
always changing and the model needs to be changed to fit the new information.In online learning,
the model can be updated with each new data point or a small batch of data, and the model is
adjusted accordingly.
One key difference between batch learning and online learning is that batch learning requires all
data to be available upfront, while online learning can work with data that arrives in a stream.
Batch learning also usually takes more time and computing resources, while online learning can
be faster and more efficient.
Whether you choose batch learning or online learning depends on the problem, the resources you
have, and the way the data is set up. Most of the time, batch learning is used when the dataset is
fixed and already known, while online learning is used when the data is dynamic and always
changing.
Main Challenges of Machine Learning.
Machine learning is a powerful tool that lets computers learn from data and then use what
they've learned to make predictions or decisions.However, machine learning is not without its
challenges. Here are some of the main challenges of machine learning:
A. Data quality: machine learning algorithms require high-quality data to learn from. If the
data is noisy, incomplete, or biased, the machine learning model will produce inaccurate
or unreliable results.
B. Overfitting: Machine learning models can sometimes become too complex and fit too
closely to the training data. This can lead to overfitting, where the model does well with
the data it was trained on but poorly with data it has never seen before.
C. Interpretability: Many machine learning models are difficult to interpret, making it hard to
understand how they are making predictions or decisions. This can be a problem in
applications where transparency and accountability are important.
D. Bias: Machine learning models can be biased if the training data is biased or if the algorithm
itself is biased. This can result in unfair or discriminatory outcomes.
E. Scalability: Machine learning models can be hard to train and deploy because they take a
lot of time and resources.This can be a challenge for applications that require real-time or
near-real-time predictions.
F. Security and privacy: Machine learning models can be attacked, like with adversarial
attacks, where the attacker changes the input data on purpose to make the model give wrong
results.Also, machine learning models can sometimes reveal private information about
people, which raises concerns about privacy.
G. Domain knowledge: machine learning algorithms require a good understanding of the
domain and the problem at hand. Without this knowledge, it can be difficult to choose
appropriate features, preprocess the data, and evaluate the performance of the model.