Module

Description

Mathematical Foundations

for

Machine Learning

Linear Algebra:

·       Vectors, Matrices, Matrix Multiplication

·       Dot Product and Eigenvalues

Calculus:

·       Derivatives, Chain Rule

·       Gradient Descent basics

Probability and Statistics:

·       Probability Distributions (Normal, Binomial)

·       Mean, Median, Variance, Standard Deviation

·       Hypothesis Testing and p-values

Python Programming

Python Basics:

·       Data types, Loops, Functions

·       List Comprehensions

Python Libraries:

·       Numpy, Pandas for data handling

·       Matplotlib, Seaborn for visualization

Basic Data Handling:

·       Loading and exploring datasets

·       Handling missing values and duplicates

·       Data normalization and standardization

Introduction to

Machine Learning

Definition and Applications of Machine Learning

Types of ML:

·       Supervised: Regression, Classification

·       Unsupervised: Clustering, Dimensionality Reduction

 ML Workflow:

·       Problem Definition

·       Data Collection and Preparation

·       Model Training and Evaluation

Supervised Learning

Regression:

·       Simple Linear Regression

·       Multiple Linear Regression

·       Polynomial Regression

·       Evaluation Metrics:

o   Mean Squared Error (MSE)

o   R² Score

Classification:

·       Logistic Regression

·       K-Nearest Neighbors (KNN)

·       Decision Trees

·       Evaluation Metrics:

o   Confusion Matrix

o   Precision, Recall, F1-Score

o   ROC-AUC

Unsupervised Learning

Clustering:

·       K-Means Clustering

·       Hierarchical Clustering

·       DBSCAN

Dimensionality Reduction:

·       Principal Component Analysis (PCA)

·       t-SNE for visualization

Feature Engineering

Handling missing data

Encoding categorical variables

·       One-Hot Encoding, Label Encoding

Feature Scaling:

·       Normalization, Standardization

Feature Selection Techniques

Feature Extraction (e.g., TF-IDF for text data)

Model Evaluation and Improvement

Train-Test Split

Cross-Validation

Hyperparameter Tuning: