aNet.ai · NewUU course · Dr. Roy’s lectures
Introduction to Machine Learning
From mathematical modeling and linear regression to neural networks, GANs and federated learning. 18 topics, 151 lessons, 84 formulas and 107 practice questions — earn XP, keep your streak, and level up as you go.
Start learning · Mathematical modeling ~21 h of study151 lessons to goTopic 1: Introduction to Machine Learning
0%complete
Daily goal
0/ 40 XP
Finish a lesson or a quiz round to start today.
Streak & level
0day streak
0best streak
1
Data point50 XP to level 2
Spaced review
0questions due now
107 questions across every topic. Answer correctly to push them further out.
Start a roundInsight of the day
Achievements
0/23- First steps — Complete your first lesson
- Warming up — Complete 10 lessons
- Deep diver — Complete 50 lessons
Your activity
0 active days · 0 XP totalLess More
Foundations
0/62 lessons- 1Introduction to Machine LearningWhat ML is, how it relates to AI and deep learning, and the vocabulary used for the rest of the course.
- 2Mathematical FoundationsThe linear algebra and probability used by every model in the course.
- 3Learning & GeneralizationWhy doing well on training data is not enough, and how to build models that work on new data.
- 4Learning as an Optimization ProblemTraining as minimising a loss: risk, gradients and the family of gradient-descent methods.
- 5Similarity / Divergence MeasuresMeasuring how different two probability distributions are — the core of generative modelling.
Supervised learning
0/42 lessons- 6Linear RegressionThe first complete model: a linear function fit by least squares, in closed form or by gradient descent.
- 7ClassificationPredicting categories with probabilities: logistic regression, its boundary and its multiclass extension.
- 8Model EvaluationMeasuring classifier quality honestly and comparing models fairly.
- 9Overfitting and RegularizationRevisiting the bias–variance trade-off with the tools to control it.
- 10Decision Trees & Ensemble LearningRule-based trees, and combining many of them into forests and boosted ensembles.
- 11Support Vector MachinesClassifiers that maximise the margin, and kernels that bend the boundary.
- 12Nearest-Neighbor & Probabilistic MethodsPredicting from similar examples, and classifying with Bayes' theorem.
Unsupervised learning
0/14 lessonsNeural & generative models
0/29 lessons- 15Introductory Neural NetworksFrom a single neuron to multilayer networks trained with backpropagation, plus convolutional networks.
- 16Advanced TopicsLearning to generate data: GANs as a two-player game over a divergence, and Boltzmann machines.
- 17Distributed MLTraining across many machines or devices, including when the data cannot leave them.