MACHINE LEARNING
UNIT I: BASICS OF MACHINE LEARNING
1.1 Introduction to Learning Problems
• What is Learning?
• Components of a Learning Problem
• Perspectives and Issues in Machine Learning
1.2 Concept Learning
• Definition and Scope
• Version Spaces
• Candidate Elimination Algorithm
1.3 Inductive Bias
• Types of Inductive Bias
• Importance in Learning
1.4 Decision Tree Learning
• Representation of Decision Trees
• ID3 Algorithm
• Heuristic-Based Space Search
• Attribute Selection Measures (Information Gain, Gini
Index)
UNIT II: NEURAL NETWORKS AND GENETIC
ALGORITHMS
2.1 Neural Networks
• Representation of Neural Networks
• Perceptrons and Linear Separability
• Multilayer Perceptrons (MLPs)
• Backpropagation Algorithm
• Activation Functions
• Training and Convergence
2.2 Advanced Topics in Neural Networks
• Limitations
• Deep Learning Concepts (brief overview)
2.3 Genetic Algorithms
• Biological Motivation
• Hypothesis Space Search
• Genetic Operators: Selection, Crossover, Mutation
• Fitness Function
2.4 Genetic Programming
• Evolutionary Computation
• Syntax Trees
2.5 Models of Evolution and Learning
• Lamarckian vs Darwinian Models
• Applications of Genetic Learning
UNIT III: BAYESIAN AND COMPUTATIONAL LEARNING
3.1 Bayesian Concepts
• Bayes Theorem
• Concept Learning using Bayes
• Maximum Likelihood Estimation (MLE)
• Minimum Description Length (MDL) Principle
3.2 Bayesian Algorithms
• Bayes Optimal Classifier
• Gibbs Algorithm
• Naïve Bayes Classifier
• Bayesian Belief Networks
3.3 Expectation Maximization (EM) Algorithm
• Use Cases
• Steps of EM
3.4 Probability-Based Learning
• Probabilistic Models
• Learning from Noisy Data
3.5 Computational Learning Theory
• Sample Complexity
• PAC Learning
• Mistake Bound Model
• Finite vs Infinite Hypothesis Spaces
UNIT IV: INSTANCE-BASED LEARNING
4.1 Fundamentals of Instance-Based Learning
• Lazy vs Eager Learning
• Characteristics and Applications
4.2 K-Nearest Neighbors (K-NN)
• Algorithm
• Distance Metrics (Euclidean, Manhattan)
• Effect of K on Accuracy
4.3 Locally Weighted Regression
• Concept
• Applications and Weighting Functions
4.4 Radial Basis Functions (RBF)
• Kernel Functions
• Use in Function Approximation
4.5 Case-Based Reasoning (CBR)
• CBR Cycle: Retrieve, Reuse, Revise, Retain
• Examples and Use Cases
UNIT V: ADVANCED LEARNING TECHNIQUES
5.1 Learning Sets of Rules
• Rule-Based Learning Systems
• Sequential Covering Algorithm
5.2 First-Order Rule Learning
• First-Order Logic Representations
• Rule Sets vs Decision Trees
5.3 Induction on Inverted Deduction
• Logical Representations
• Inverting Resolution
5.4 Analytical Learning
• Using Prior Knowledge
• Perfect Domain Theories
• Explanation-Based Learning (EBL)
5.5 FOCL Algorithm
• First-Order Combined Learner
• Integration of EBL and Empirical Methods
5.6 Reinforcement Learning
• Key Concepts: Agent, Environment, Reward
• Exploration vs Exploitation
• Temporal Difference Learning
• Q-Learning and SARSA (overview)