Tags: [[study]], [[ai]]

Stanford CS230 ｜ Autumn 2025 ｜ Lecture 2 Supervised, Self-Supervised, & Weakly Supervised Learning

Date: 2026-01-24 20:31 Source: https://www.youtube.com/watch?v=DNCn1BpCAUY

Notes (상세 내용)

  • [0:00:15] 핵심 개념 1: …
  • [0:01:40] 예시: …
  • [0:03:22] 핵심 개념 2: …
  • 01:10: Today’s lecture Goal
    • Better way to make decisions in AI projects
  • Later Classes you will see
    • Adversarial attacks and defences
    • Deep Reinforcement Learning
    • Retrieval Augmented Generation
    • AI Agents
    • Multiagent System
    • Neural Networks
  • 02:11: Today’s outline
    • I. Recap’ of the week
    • Il. Supervised Learning Projects
      1. Day & Night Classification
      2. Trigger Word Detection
      3. Face Verification
    • III. Self-Supervised Learning & Weakly Supervised Learning Projects
      1. Image Embeddings
      2. Multi-Modal Embeddings
    • IV. (If time allows) Adversarial Attacks
  • 03:30: I. Recap’ of the week
    • Input: Cat Image
    • Output: 0 or 1
    • Model = Architecture + Parameters
    • How does the model learn?
      • Gradient Descent Optimization
        • Use a loss function: compare the ground truth
        • If 0, get penalty in order to give feedback to the parameters
      • Repeat the parameter updates
    • ![[mx-img-yqxomstoq41rll4rdbpdtjiz-pt5m36_77s.jpg Stanford CS230 ｜ Autumn 2025 ｜ Lecture 2 Supervised, Self-Supervised, & Weakly Supervised Learning - 05:36 50]] 05:36

Cue (질문/키워드)

[!cue] 핵심 질문/키워드 1

  • Timestamp: 00:00
  • Note: 여기에 강의 내용을 상세히 기록합니다.
  • My Thought: (내 생각/연결 아이디어)

[!cue] 핵심 질문/키워드 2

  • Timestamp: 05:30
  • Note: 두 번째 핵심 내용입니다.
  • My Thought: …

Summary (내 언어로 요약)

[!summary] Summary 이 비디오는 [주제]에 대한 [핵심 내용]을 다루고 있다. 특히 [가장 중요한 내용]을 강조하며, 이는 [결론]으로 이어진다. 이 내용을 통해 나는 [나의 깨달음]을 얻었다.

3줄 요약

Review Questions (복습용)

  • (나중에 이 칸을 채우거나 AI에게 채워달라고 요청)