Machine Learning Concepts
Learning paradigms, tasks and the workflow of building a model.
This is an overview — pick a sub-topic below.
Sub-topics (24)
Artificial Intelligence
Computer systems that perform tasks normally needing human intelligence — perception, reasoning, language and decision-making.
Machine Learning
Algorithms that improve at a task by learning patterns from data.
Supervised Learning
Learn from labelled examples (input → known answer) to predict answers for new inputs.
Unsupervised Learning
Find structure in unlabelled data — clusters, anomalies, compressed representations.
Semi-Supervised and Self-Supervised Learning
Use a little labelled data plus lots of unlabelled data; self-supervised models create their own labels (e.g. predict the next word).
Reinforcement Learning
An agent learns by trial and error, maximising reward from an environment.
Deep Learning
Machine learning with many-layered neural networks that learn features directly from raw data.
Transfer Learning
Start from a model pretrained on a big dataset and fine-tune it for your task with little data.
Federated, Online, Active and Continual Learning
Train across devices without sharing raw data (federated), update as data streams in (online), ask humans to label the most useful examples (active), and keep learning without forgetting (continual).
Ensemble Learning
Combine many models (bagging, boosting, stacking) for better accuracy.
Classification and Regression
Predict a category or a continuous value.
Clustering
Group similar items without labels: k-means, DBSCAN, hierarchical clustering.
Dimensionality Reduction
Compress many features into fewer while keeping structure: PCA, t-SNE, UMAP.
Feature Selection and Extraction
Keep the most useful inputs, or compute better ones from raw data.
Anomaly Detection
Find rare, unusual patterns — machine faults, fraud, intrusion.
Recommendation Systems
Suggest items users may like via collaborative filtering, content similarity or hybrids.
Model Evaluation
Measure performance honestly: train/validation/test splits, cross-validation and the right metric.
Hyperparameter Tuning
Search for the best settings (learning rate, depth, …) with grid, random or Bayesian search.
Model Interpretability and Explainable AI
Understand why a model predicts what it does — feature importance, SHAP values, saliency maps.
AutoML
Automatically try models, features and hyperparameters to find a good pipeline.
Bayesian Machine Learning and Causal Inference
Model uncertainty with probability distributions; estimate cause-and-effect, not just correlation.
Meta-, Few-Shot and Zero-Shot Learning
Learn to learn new tasks from very few (or zero) labelled examples.
Multimodal and Neuro-Symbolic AI
Models that combine vision, text and audio; and systems mixing neural networks with symbolic reasoning.
Evolutionary Algorithms, Genetic Algorithms and Fuzzy Logic
Optimise by simulated evolution (selection, crossover, mutation); reason with degrees of truth.
Spotted a mistake or want to add detail? Suggest an improvement — Content Contributors' edits are reviewed by editors before they go live.
Discussion (0)
Sign in or join free to ask questions and discuss this topic.