Machine Learning

Pre-requisites, Topics and outcomes

Course Name: ME (Master of Engineering (CSE)), I year, 2024. Institute name: MBM University, Jodhpur

Pre-requisites: Mathematics, Statistics, Probability Theory, Computer Fundamentals, Python, AI

Syllabus

Basic Concepts

Basics, learning in animals, supervised learning, unsupervised learning, Reinforcement learning, Neural Network based learning, Deep learning Networks

Slide 1: Introduction Introduction to ML, Machine Learning defined, example of application in spam filtering, Problem with baits shyness learning, criteria about theory of ML, justification for machine learning, why hand coded system are not enough?
Slide 2: Supervised and Unsupervised Learning Supervised Learning, Its formal definition, unsupervised learning, unsupervised algorithm examples
Slide 3: Reinforcement learning, neural network based learning, Deep Learning Data for supervised and unsupervised Learning, Reinforcement learning, learning through neural networks, Deep learning Networks
HW #1: Basic Concepts

Advanced Concepts

Reinforcement learning, neural networks, components of a neuron, ANN, Naive Bayes, KNN, classifiers ...

Slide 4: RL and NN Reinforcement learning: policy, reward function, mapping, state-space, task example, function in RL. Simple architecture of NN, properties of neuron, its applications. Basic model of artificial neuron, its connections, and math. expression.
Slide 5: Basic Algorithms: Naive Bayes & KNN Labeling email as spam/nospam, Vector space, applying Naive Bayes, Nearest neighbor classifier, Euclidean distance.
Slide 6: Training sets and Classifiers Bayes theorem, training set, Classifier, single attribute case, calculating conditional probabilities.
Slide 7: Naive Bayes Classifiers Discrete attribute vectors, computing vectors probability, training set, naive Bayes Classifier.
Slide 8: Bayesian classifier for continuous attributes Continues attribute vectors, Bayes formula for continuous attribute vector, parameter values.
Slide 9: Linear classifier Linear classifier, learning from examples, perceptron learning.