Seven Ideas in Machine Learning
Visual explanations of ML, with a two-minute history to set the scene.
IIT Gandhinagar
Teaching
25 course offerings, 562 videos and 86 cheatsheets since 2018.
Search results
This semester · Aug 2026
Deep Learning
ES 667. Losses, gradients, optimization, regularization and attention.
Courses
2018–2026One row per course. Each chip is one offering and links to that semester’s course site.
| Course | Taught | Years | Offerings |
|---|---|---|---|
| Deep LearningES 667Cheatsheets · Aug 2026 | 1× | 2026 | |
| Software Tools and Techniques for AICS 203Cheatsheets · Jan 2026 | 1× | 2026 | |
| Principles of AIES 119 | 1× | 2026 | |
| Machine LearningES 654 / ES 335Cheatsheets · Aug 2025 | 8× | 2019–2025 | |
| AI for Social GoodCS 691 | 1× | 2025 | |
| Probability, Statistics, and Data VisualizationES 114Cheatsheets · Jan 2025 | 1× | 2025 | |
| Probabilistic Machine LearningCS 691 / ES 661Cheatsheets · Aug 2023 | 2× | 2022–2023 | |
| ComputingES 112 | 3× | 2020–2022 | |
| Ubiquitous ComputingCS 691 | 1× | 2021 | |
| Operating SystemsCS 301 | 3× | 2018–2020 | |
| Graduate SystemsCS 612 | 3× | 2018–2020 |
Aug 2026 teaching now Jan 2024 lectures recorded Jan 2023 course site § cheatsheets
Videos
562 videos in 19 playlistsVisual explanations
8 videosOne-minute lessons
270 videosThree-minute lessons
109 videosCourse recordings
158 videosDeep learning · 2026
Full lectures from ES 667.
Software tools for AI · 2026
Machine learning · 2024
Probabilistic ML · 2023
Machine learning · 2020
Operating systems · 2018
AI for social good · 2025
Guest talks from the January course.
Science & everyday systems
2 videosIndian scientists
Anna Mani and the measurement of the atmosphere.
How things work
Animated explanations of the systems we use every day.
Practical tutorials
17 videosCheatsheets
86 PDFsTwo-page references from my courses. Open a course to see its sheets, or search for a topic above.
Deep learningES667 · Aug 2026Losses, gradients, optimization, regularization and attention.12 PDFs
- Where Deep Learning Losses Come From IPDF · 2 pages
- Where Deep Learning Losses Come From IIPDF · 2 pages
- Likelihood to Loss in PracticePDF · 2 pages
- From Linear Models to Neural NetworksPDF · 2 pages
- Optimization I — Gradient Estimates & Step SizesPDF · 2 pages
- Backpropagation & Autograd from ScratchPDF · 2 pages
- Optimization II — Momentum, RMSProp & AdamPDF · 2 pages
- Vector, Matrix & Affine AutogradPDF · 2 pages
- Valid Outputs by Design & Learning-Rate SchedulesPDF · 2 pages
- Generalization & RegularizationPDF · 2 pages
- From characters to next-token predictionPDF · 2 pages
- Self-attention, from first principlesPDF · 2 pages
Software tools for AICS203 · Jan 2026Data, APIs, Git, experiments, deployment and AI agents.15 PDFs
- Collecting data over HTTP and APIsPDF · 2 pages
- Scraping web pagesPDF · 2 pages
- Validating data before trainingPDF · 2 pages
- Labelling dataPDF · 2 pages
- Labelling less: active learning, weak supervision, LLM labelsPDF · 2 pages
- Data augmentationPDF · 2 pages
- LLM APIs and multimodal modelsPDF · 2 pages
- Evaluating models honestlyPDF · 2 pages
- Tuning and experiment trackingPDF · 2 pages
- Reproducible runs: seeds, environments, DockerPDF · 2 pages
- Data drift and monitoringPDF · 2 pages
- Profiling and quantizationPDF · 2 pages
- Building AI agents from scratchPDF · 2 pages
- From script to web app and APIPDF · 2 pages
- Git for ML projectsPDF · 2 pages
Machine learningES335 · Aug 2025Trees, regression, neural networks, clustering and support vector machines.28 PDFs
- Conventions and metricsPDF · 2 pages
- Decision trees I: entropy and information gainPDF · 2 pages
- Decision trees II: real inputs, real outputs, depthPDF · 2 pages
- Bias and variancePDF · 2 pages
- Cross-validation and hyperparametersPDF · 2 pages
- Ensembles: bagging, random forests, boostingPDF · 2 pages
- Maths for ML: gradients, rank, contoursPDF · 2 pages
- Linear regression I: normal equation and geometryPDF · 2 pages
- Linear regression II: basis functions, collinearity, dummiesPDF · 2 pages
- Gradient descent and SGDPDF · 2 pages
- Matrix factorisation for recommendationsPDF · 2 pages
- Ridge regressionPDF · 2 pages
- Lasso, subgradients and coordinate descentPDF · 2 pages
- Constrained optimisation and KKTPDF · 2 pages
- Logistic regressionPDF · 2 pages
- Convexity, Hessians and Newton's method (IRLS)PDF · 2 pages
- Precision-recall curvesPDF · 2 pages
- Multi-layer perceptronsPDF · 2 pages
- Automatic differentiationPDF · 2 pages
- Next-character predictionPDF · 2 pages
- Convolutional networksPDF · 2 pages
- Unsupervised learning: clustering and PCAPDF · 2 pages
- Reinforcement learningPDF · 2 pages
- K-nearest neighbours and approximate searchPDF · 2 pages
- Support vector machines I: maximum marginPDF · 2 pages
- Support vector machines II: kernels and soft marginPDF · 2 pages
- Naive Bayes and Bayesian networksPDF · 2 pages
- Time-series forecasting and sequence modelsPDF · 2 pages
Probability, statistics & data visualizationES114 · Jan 2025Python, probability, distributions, expectation and data visualization.15 PDFs
- Arrays and tables: NumPy and PandasPDF · 2 pages
- Plotting and honest chartsPDF · 2 pages
- Sets, sample spaces and eventsPDF · 2 pages
- Counting and the complement rulePDF · 2 pages
- Conditional probability, independence and Bayes' rulePDF · 2 pages
- Random variables and PMFsPDF · 2 pages
- Discrete distributionsPDF · 2 pages
- Expectation and variancePDF · 2 pages
- Continuous distributions and densitiesPDF · 2 pages
- CDFs and samplingPDF · 2 pages
- Joint distributionsPDF · 2 pages
- Covariance, correlation and the bivariate normalPDF · 2 pages
- Sums of random variables, LLN and CLTPDF · 2 pages
- Random vectors and PCAPDF · 2 pages
- Likelihood and log probabilitiesPDF · 2 pages
Probabilistic machine learningES661 · Aug 2023Bayesian models, sampling, variational inference and uncertainty.16 PDFs
- Probability refresher for MLPDF · 2 pages
- Maximum likelihood estimationPDF · 2 pages
- MLE for linear and logistic regressionPDF · 2 pages
- MAP estimation and priorsPDF · 2 pages
- Multivariate normal and Bayesian linear regressionPDF · 2 pages
- Bayesian logistic regressionPDF · 2 pages
- Laplace approximationPDF · 2 pages
- Sampling I: from uniform to any distributionPDF · 2 pages
- Neural processesPDF · 2 pages
- Sampling II: rejection and importance samplingPDF · 2 pages
- Markov chain Monte CarloPDF · 2 pages
- Hidden Markov modelsPDF · 2 pages
- Active learningPDF · 2 pages
- Bayesian optimisationPDF · 2 pages
- Information theoryPDF · 2 pages
- Variational inferencePDF · 2 pages