ES 335 IIT Gandhinagar

Home Lectures Assessment FAQ Recordings

Lectures

Each lecture lists its slides and reading first, then its full class recording, then practice notebooks. Slides and notebooks come from ml-teaching, which also lists all slides and all notebooks.

SlidesFull recordingNotebookReading

Foundations1–2Trees, model fit and ensembles3–7Linear models and optimisation8–17Classification and neural networks18–26Neighbours, unsupervised and reinforcement learning27–30Support vector machines31–35

Lecture recordings playlist

Foundations

Lectures 1–2

·	Prerequisite test An open-book test on Python, linear algebra, probability and calculus, required to take the course.	Prerequisite test
Lecture 01	Introduction and logistics	Slides Recording Transcription and local LLMsSegmentation and localisationPrompt to image
Lecture 02	Conventions, metrics, classification and regression	Slides Recording Rule-based vs MLConfusion matrixDummy baselines

Trees, model fit and ensembles

Lectures 3–7

Lecture 03	Decision trees I	Slides Recording EntropyDecision trees: real outputDecision trees: real input, discrete output
Lecture 04	Decision trees II, bias and variance	Slides Recording Bias and varianceDecision trees: overfitting and underfitting
Lecture 05	Tutorial on NumPy, pandas and classes; cross-validation	Slides: Cross-validation Tutorial recordingLecture on cross-validation NumPy, pandas, entropyClasses and treesCross-validation and tuning hyperparameters
Lecture 06	Cross-validation and ensemble methods	Slides: Ensemble methods Recording Cross-validation and tuning hyperparameters
Lecture 07	Random forests Lecture on random forests, and a tutorial on random forests and meshgrid.	Slides Recording Random forest feature importance from scratchMeshgrid

Linear models and optimisation

Lectures 8–17

Lecture 08	Linear regression	Slides Recording
Lecture 09	Linear regression: basis functions	Slides Recording Basis functions
Lecture 10	Linear regression: geometric interpretation	Slides Recording Geometric linear regression
Lecture 11	Multicollinearity and dummy variables	Slides Recording Dummy variables and multicollinearity
Lecture 12	Gradient descent I	Slides Recording Taylor seriesContour plots and gradient descent
Lecture 13	Tutorial: Google Cloud, forecasting, polynomial regression Google Cloud, forecasting, and polynomial regression with basis functions.	Reading: Polynomial regression Polynomial regression (GitHub, not rendered)Autoregressive model
Lecture 14	Gradient descent II	Slides Recording Gradient descent
Lecture 15	Movie recommendation and matrix factorisation	Slides Recording Movie recommendation
Lecture 16	Regularised linear regression: ridge and lasso	Slides: Ridge regressionSlides: Lasso regression Recording
Lecture 17	Coordinate descent, time complexity, maths for ML, convexity Not recorded.	Slides: Coordinate descentSlides: Time complexitySlides: Maths for MLSlides: Convexity scikit-learn on GPU (GitHub, not rendered)

Classification and neural networks

Lectures 18–26

Lecture 18	Guest lecture: adversarial examples for ML Guest lecture by Arjun Bhagoji.	Speaker: Arjun Bhagoji Adversarial examples tutorial (GitHub, not rendered)
Lecture 19	Logistic regression I	Slides Recording Why use logitsLogistic regression in PyTorch
Lecture 20	Logistic regression II	Slides Recording Logistic regression on circular data
Lecture 21	Multi-layer perceptrons I	Slides Recording
Lecture 22	Multi-layer perceptrons II	Slides Recording Vectorised networks with vmap (GitHub, not rendered)
Lecture 23	Next-token prediction	Slides Recording Generating names
Lecture 24	Autograd	Slides Recording Automatic differentiation
Lecture 25	Convolutional neural networks I	Slides: 1D CNN Recording 1D CNNEdge detectionLeNet
Lecture 26	Convolutional neural networks II	Recording CNN (GitHub, not rendered)

Neighbours, unsupervised and reinforcement learning

Lectures 27–30

Lecture 27	K-nearest neighbours	SlidesReading: Parametric vs non-parametric (MLSS)Reading: Machine Learning MasteryReading: StackExchangeReading: Sebastian Raschka Recording Parametric vs non-parametricCurse of dimensionality
Lecture 28	Reinforcement learning	Slides Recording Reinforcement learning IReinforcement learning II
Lecture 29	Unsupervised learning	Slides Recording PCA (GitHub, not rendered)
Lecture 30	Approximate nearest neighbours	Slides Recording KNN variants

Support vector machines

Lectures 31–35

Lecture 31	Constrained optimisation	Slides: Constrained optimisation ISlides: Constrained optimisation II (self-study) Part II recording, from an earlier run SVM with CVXOPT (GitHub, not rendered)
Lecture 32	Support vector machines I	Slides Recording SVM
Lecture 33	Support vector machines II: kernels	Slides Recording Primal and dualKernels
Lecture 34	Support vector machines III: QP, kernels as similarity	Recording Quadratic programmingUnderstanding kernels
Lecture 35	Support vector machines IV: kernels, soft margin	Slides Recording Understanding kernelsSoft margin

Notebooks marked "GitHub, not rendered" are in the ml-teaching repository only. Quizzes, mock quizzes, the mid-semester exam discussion and the final revision class are on the Assessment page.