Skip to main content

Syllabus — Machine Learning and Data Analytics

Year I, Part I — MSNCS, IOE Pulchowk, Tribhuvan University. 4 credits.

Chapter 1 — Basics of Machine Learning (7 marks)​

  • 1.1 History of machine learning, definition of learning, types of learning, importance of ML
  • 1.2 Statistics review: Min, Max, Mean, Mode, Median, Standard deviation, MSE
  • 1.3 ML terminology: class, pattern, feature, training, validation, test data
  • 1.4 Feasibility of learning — error and noise — training vs testing
  • 1.5 Generalization tradeoff — bias and variance — learning curve
  • 1.6 Overfitting and Underfitting

Chapter 2 — Data Analytics Process (9 marks)​

  • 2.1 Process of data analytics
  • 2.2 Data types and attributes
  • 2.3 Data pre-processing
  • 2.4 Visualization and exploring data
  • 2.5 Descriptive, diagnostic, predictive, prescriptive analytics
  • 2.6 Architectural design patterns and stack for handling Big Data

Chapter 3 — Supervised Learning (9 marks)​

  • 3.1 Definition and classification problem
  • 3.2 Classifiers and discriminant functions
  • 3.3 Linear supervised learning models: linear regression, Perceptron
  • 3.4 Learning neural network structures
  • 3.5 Decision tree representation model, basic decision tree algorithm, applications
  • 3.6 Support vector machines and applications

Chapter 4 — Bayesian Decision Based Learning (9 marks)​

  • 4.1 Bayes probability theory and conditional probability
  • 4.2 Decision surfaces and classifying with Bayes decision theory
  • 4.3 Bayesian belief network and applications
  • 4.4 Gradient descent method
  • 4.5 K-nearest neighbor

Chapter 5 — Unsupervised Learning and Dimensionality Reduction (9 marks)​

  • 5.1 Introduction to clustering, criterion function for clustering
  • 5.2 Algorithms for clustering: K-means, hierarchical, and other methods
  • 5.3 Dimensionality reduction techniques and need
  • 5.4 Principal component analysis (PCA)
  • 5.5 Linear discriminant analysis (LDA)

Chapter 6 — Measures for Performance Evaluation (8 marks)​

  • 6.1 Classification accuracy
  • 6.2 Confusion matrix
  • 6.3 Misclassification costs
  • 6.4 Sensitivity and specificity, recall, precision, F1-score
  • 6.5 ROC curve, box plot, confidence interval
  • 6.6 Cross-validation

Chapter 7 — Deep Learning Basics (9 marks)​

  • 7.1 Definition of deep networks
  • 7.2 Feed-Forward and backpropagation
  • 7.3 Activation functions: Sigmoid, Tanh, ReLU, Softmax
  • 7.4 Convolutional neural networks (CNN): CNN architectures
  • 7.5 Recurrent neural networks (RNN): RNN architectures
  • 7.6 ML applications in security: anomaly/intrusion detection, malware/phishing/fraud detection
· min read