Engineering & CS
10 topics
Engineering & CS
Machine learning
Machine learning gives computers the ability to learn from data without being explicitly programmed. It sits at the intersection of statistics, linear algebra, and optimisation — and shares its mathematical toolkit with almost every other quantitative field.
Topics in this field
Backpropagation
Efficient computation of gradients in neural networks via the chain rule of calculus.
Cross-Validation
A family of resampling techniques for estimating generalization error and selecting models without a separate held-out test set.
Decision Trees
Recursive partitioning of the feature space using impurity-based splitting criteria to build interpretable prediction rules.
Gaussian Processes
A non-parametric Bayesian approach that places a prior directly over functions.
Gradient Descent
Iterative optimisation by following the steepest downhill direction. The engine of modern ML.
Linear Regression
The simplest supervised learning model — mathematically identical to econometric OLS.
Neural Networks
Universal function approximators trained by backpropagation, forming the foundation of modern deep learning.
Principal Component Analysis
Linear dimensionality reduction by projecting data onto directions of maximum variance via eigendecomposition of the covariance matrix.
Random Forests
An ensemble method combining bagged decision trees with random feature subsets to produce low-variance, high-accuracy predictions.
Support Vector Machines
Maximum-margin classifiers that find the optimal separating hyperplane using convex quadratic programming and the kernel trick.