Introduction
Types of Machine Learning
Common Machine Learning
Algorithms
Steps in Machine Learning
Appliations of Machine
Learning
Conclusion
Subfield of AI: Enables machines to mimic intelligent
behavior.
Defined by Arthur Samuel: "Computers learn without explicit
programming."
Adapts Automatically: Models improve with new data, unlike
traditional programming.
Robotic process automation Sales optimization Customer service
Combining RPA and ML ML analyzes customer
Chatbots and virtual
automates complex tasks like data for sentiment
assistants automate routine
mortgage processing. analysis, sales forecasting,
tasks, speeding up
and churn prediction.
resolutions.
Security Digital Marketing Fraud Prevention
ML enhances threat Identifies new Real-time detection of
analysis and customers and targets suspicious activity in
improves responses them with personalized
financial transactions.
to cyberattacks. marketing.
01 Features
Features are fields used as input.
A feature means the property of
your training data.
01 Features 02 Label
The output you get from your
Features are fields used as input.
model after training is called
A feature means the property of
label. It is the thing we are
your training data.
predicting.
01 Features 02 Label
Features are fields used as input. The output you get from your
A feature means the property of model after training is called
your training data. label. It is the thing we are
predicting.
01 Features 02 Label
Features are fields used as input. The output you get from your
A feature means the property of model after training is called
your training data. label. It is the thing we are
predicting.
01
Supervised learning involves training a model on
a labeled dataset, where each data point has a
corresponding label. The model learns to map
inputs (features) to outputs (labels) and can
make predictions on new, unseen data.
01
Supervised learning involves training a model on
a labeled dataset, where each data point has a
corresponding label. The model learns to map
inputs (features) to outputs (labels) and can
make predictions on new, unseen data.
01
Supervised learning involves training a model on
a labeled dataset, where each data point has a
corresponding label. The model learns to map
inputs (features) to outputs (labels) and can
make predictions on new, unseen data.
01
Supervised learning involves training a model on
a labeled dataset, where each data point has a
corresponding label. The model learns to map
inputs (features) to outputs (labels) and can
make predictions on new, unseen data.
01
Supervised learning involves training a model on
a labeled dataset, where each data point has a
corresponding label. The model learns to map
inputs (features) to outputs (labels) and can
make predictions on new, unseen data.
01
01
Unsupervised learning involves training a model
on unlabeled data, allowing it to find patterns
and relationships on its own. Common tasks
include clustering (grouping similar data) and
dimensionality reduction (simplifying data by
reducing variables).
01
Unsupervised learning involves training a model
on unlabeled data, allowing it to find patterns
and relationships on its own. Common tasks
include clustering (grouping similar data) and
dimensionality reduction (simplifying data by
reducing variables).
01
Unsupervised learning involves training a model
on unlabeled data, allowing it to find patterns
and relationships on its own. Common tasks
include clustering (grouping similar data) and
dimensionality reduction (simplifying data by
reducing variables).
01
Unsupervised learning involves training a model
on unlabeled data, allowing it to find patterns
and relationships on its own. Common tasks
include clustering (grouping similar data) and
dimensionality reduction (simplifying data by
reducing variables).
01
01
01
Reinforcement learning is a type of machine
learning where an agent learns by interacting
with its environment. It receives rewards or
penalties for its actions, aiming to maximize its
total reward
01
Reinforcement learning is a type of machine
learning where an agent learns by interacting
with its environment. It receives rewards or
penalties for its actions, aiming to maximize its
total reward
01
Reinforcement learning is a type of machine
learning where an agent learns by interacting
with its environment. It receives rewards or
penalties for its actions, aiming to maximize its
total reward.
01
Reinforcement learning is a type of machine
learning where an agent learns by interacting
with its environment. It receives rewards or
penalties for its actions, aiming to maximize its
total reward.
01
A machine learning algorithm functions like a recipe that guides computers to learn from data and
make predictions based on input variables. By analyzing large datasets, these algorithms identify
patterns and insights autonomously.
One of the most popular supervised learning algorithms for classification tasks.
Splits data into branches based on decision rules, leading to a final prediction.ivides the data into
homogeneous sets based on significant attributes.
Modeled as a binary tree.Each node represents an input
variable (x) and a split point on that variable.
Leaf nodes contain the output variable (y) to make
predictions.
Predictions are made by traversing the tree until reaching a
leaf node.
Random Forest: Combines multiple decision trees to improve accuracy and reduce
overfitting.
Builds multiple decision trees using randomly chosen subsets of data and features.
Adds randomness by selecting suboptimal split points to create diverse trees.
By using multiple trees, random forests
reduce the risk of overfitting and provide
more accurate and stable predictions than
a single decision tree.
The bootstrap is a powerful statistical method for estimating a quantity from a data sample. Such as a
mean. You take lots of samples of your data, calculate the mean, then average all of your mean values to
give you a better estimation of the true mean value.
In bagging, the same approach is used, but instead for estimating entire statistical models, most
commonly decision trees. Multiple samples of your training data are taken then models are constructed
for each data sample. When you need to make a prediction for new data, each model makes a prediction
and the predictions are averaged to give a better estimate of the true output value.
Powerful and widely used machine learning classifier.
Ideal for datasets with both linear and non-linear separations.
Identifies a hyperplane that separates classes with maximum margin.
Handles outliers and overlapping data points robustly.
Effective in high-dimensional spaces.
Works well with non-linear data using the kernel trick.
Less prone to overfitting, especially on small datasets.
A line (or plane in higher dimensions) that separates the input variable space by class.
Chosen to best distinguish points belonging to different classes (e.g., class 0 vs. class 1).
Margin:
The distance between the hyperplane and the closest data points from each class.
A larger margin improves the model's robustness and generalization.
Classification algorithms generally take past data (data for which you have known outcomes),
train the model, take new data once the model is trained, ingest it, and create predictions (e.g., is
it a truck or is it a car?).
The goal of this example is to classify cricket players into batsmen or bowlers
using the runs-to-wicket ratio. A player with more runs would be considered a
batsman and a player with more wickets would be considered a bowler.
Before separating anything using high-level mathematics, let’s look at an unknown value, which
is new data being introduced into the dataset without a predesignated classification.
The next step is to draw a decision boundary, or a line separating the two classes to help classify
the new data points.
