MACHINE LEARNING

Course Duration: 12–20 Weeks (Flexible) | Mode: Offline/Online

Goal: Students become Machine Learning Engineers with strong mathematical, coding, and model-building foundations.

MODULE 1: Introduction to ML (Week 1)

What is Machine Learning?
AI vs ML vs DL
ML workflow
Types of ML: Supervised, Unsupervised, Reinforcement
Real-world ML applications
ML project lifecycle

MODULE 2: Python Essentials for ML (Week 1–2)

Data types, loops, functions
NumPy: arrays, reshaping, broadcasting
Pandas: DataFrame operations, missing values
Matplotlib & Seaborn basics

MODULE 3: Statistics & Mathematics for ML (Week 2–4)

Statistics

Descriptive stats
Probability basics
Distributions
Hypothesis testing
p-value, t-test, chi-square

Linear Algebra

Vectors, matrices
Matrix transformations
Eigenvalues & eigenvectors (overview)

Calculus for ML

Derivatives
Gradient descent intuition

MODULE 4: Data Preprocessing & EDA (Week 4–5)

Handling missing data
Outlier detection (IQR, Z-score)
Feature scaling (standardization, normalization)
Encoding categorical features
Feature engineering
Exploratory Data Analysis

MODULE 5: Supervised Learning – Regression (Week 5–6)

Topics

Linear Regression (simple & multiple)
Cost function & gradient descent
Polynomial Regression
Regularization (L1, L2, ElasticNet)

Model Evaluation

RMSE, MAE, R²
Train-test split
Cross-validation

MODULE 6: Supervised Learning – Classification (Week 6–7)

Topics

Logistic Regression
KNN
Naive Bayes
Decision Trees
Random Forest
Gradient Boosting
XGBoost / LightGBM / CatBoost (optional advanced)

Model Evaluation

Confusion matrix
Precision, Recall, F1 Score
ROC-AUC

MODULE 8: Feature Engineering & Model Improvement (Week 8–9)

Pipeline creation (Sklearn Pipeline)
Scaling & encoding pipelines
Feature selection
Hyperparameter tuning (Random Search, Grid Search)
Ensemble methods & stacking models

MODULE 9: Time Series Forecasting (Optional — Week 9–10)

Time series components
Stationarity
AR, MA, ARIMA, SARIMA
Prophet (Facebook)

MODULE 10: Introduction to Deep Learning (Week 10–12)

Neural Networks basics
Perceptron
Forward & backward propagation
Activation functions
Loss function
Optimizers
Overfitting & regularization
Keras/TensorFlow basics

MODULE 11: Computer Vision (Optional — Week 12–13)

Convolutional Neural Networks (CNNs)
Pooling, filters, feature maps
Transfer learning (VGG, ResNet, MobileNet)

MODULE 12: Natural Language Processing (Week 13–14)

Text preprocessing
Tokenization
Stemming & Lemmatization
Bag of Words
TF-IDF
Word embeddings (Word2Vec intro)