Machine Learning (BIT421CO)

Theoretical concepts and practical implementations of supervised regression/classification, decision trees, model tuning, text mining, and deep neural networks in Python.

1. Unit 1: Introduction to Machine Learning5 hrs

   • Components of learning, Learning Models: Geometric, Probabilistic, Logical models
   • Introduction to machine learning frameworks: Supervised, Unsupervised, Reinforcement Learning

2. Unit 2: Supervised Learning12 hrs

   • Linear regression, Polynomial regression
   • Logistic regression, Support Vector Machines (SVM)
   • k-NN (k-Nearest Neighbors)
   • Decision tree: Representation, ID3, C4.5, Inductive bias

3. Unit 3: Unsupervised Learning4 hrs

   • Clustering fundamentals and similarity metrics
   • k-means clustering algorithm and variants
   • k-modes algorithm for categorical data

4. Unit 4: Model Diagnosis and Tuning7 hrs

   • Evaluating a hypothesis: Bias-variance tradeoff
   • Model selection: k-fold Cross-validation
   • Ensemble methods: Random forests and bagging

5. Unit 5: Text Mining6 hrs

   • Text preprocessing: Tokenization, Stopwords removal, Stemming, Lemmatization
   • Feature representation: Bag of Words, TF-IDF
   • Text exploration and classification

6. Unit 6: Deep Learning11 hrs

   • Feedforward neural networks, Perceptron and Multilayer Perceptron
   • Cost functions, Gradient descent and Backpropagation training
   • Convolutional Neural Networks (CNNs)
   • Recurrent Neural Networks (RNNs)