machine learning

Machine learning is a subfield of artificial intelligence (AI) that builds models which improve their performance on a task by learning patterns from data rather than following hand-written rules.

Core ingredients include a representation of inputs, a loss function that scores errors, and an optimization procedure that fits parameters during training, followed by evaluation and deployment for inference.

Major paradigms are supervised learning for labeled prediction, unsupervised and self-supervised learning for structure and representations, and reinforcement learning for decision making through rewards.

Practical concerns center on generalization, overfitting, regularization, data quality and splits, metrics, and monitoring models after deployment.