# Stanford CS229 Machine Learning | Spring 2026 | Lecture 1: Introduction

## Executive summary

This lecture provides a high-level introduction to Machine Learning fundamentals (Supervised, Unsupervised, and Reinforcement Learning) within the context of modern AI. The course emphasizes understanding the mathematical foundations and core techniques behind algorithms rather than focusing on programming implementation. Key topics include model training using large datasets, advanced concepts like embeddings, and the architectural differences between traditional ML tasks and general-purpose Large Language Models (LLMs). A critical focus for system builders is placed on the necessity of optimizing ML systems for hardware compatibility and speed.

## Key takeaways

- ML Paradigm Shift: The field has moved from specific, task-oriented models to general-purpose agents (LLMs). While traditional methods still apply, the focus is on tuning fundamental model capabilities rather than building complex data pipelines for every single use case.
- The Importance of ML Systems: A planned lecture will cover 'ML system,' which addresses the critical need to make software and hardware compatible. Optimizing algorithms for speed (e.g., making them run 2x faster) is crucial due to the high cost associated with AI computation.
- Reinforcement Learning (RL) in LLMs: RL can be used to train models dealing with stochastic sampling, such as text generation. Techniques like Policy Gradient and using human feedback (e.g., RLHF/RAG) are necessary because the generation process is not differentiable.

## Technical details

- ML Definitions: Machine learning is defined as giving a computer the ability to learn without being explicitly programmed. Tom Mitchell's definition states that a program learns from experiences (E) with respect to tasks (T) and performance measure (P) if its performance improves with experience.
- Regression vs Classification: If the output is a continuous variable (e.g., price), it is Regression. If the output is a discrete label (e.g., house/townhouse, or next token from 50k options), it is Classification.
- Supervised Fine-Tuning (SFT): SFT is a form of supervised learning used in building agents. The process involves training models on labeled data pairs (X, Y).
- Latent Semantic Analysis (LSA): A technique for unsupervised learning that represents words or documents as numerical vectors. By reordering these vectors, similar concepts and documents can be grouped together.
- Diffusion Models: Used for generating realistic images from noise or instructions (e.g., text prompts). This process is inherently unsupervised as it does not rely on explicit labels.

## Practical implications

- Understanding the mathematical foundations (linear algebra, probability) is essential for building and tuning ML models.
- The ability to optimize algorithms for hardware compatibility and speed is a major engineering concern in modern AI deployment.
- LLMs fundamentally operate as massive classification problems (predicting the next discrete token), making understanding classification theory critical.

## Topics

Machine Learning Fundamentals, Supervised Learning (SFT), Unsupervised Learning (Clustering, LSA), Reinforcement Learning (RL), Large Language Models (LLMs), Diffusion Models, ML System Optimization, CS229 Machine Learning, Stanford AI Programs Info

Source: https://www.youtube.com/watch?v=DATnpGoGhM8