ML lec 01 - 기본적인 ML용어와 개념 정리

ML

• explicit programming
• spam filter : many rules
• Automatic driving : too many rules
• ML : Field of study that gives computers the ability to learn without being explicitly programmed

 

Supervised/Unsupervised learning

• supervised : learning with labeled examples (training set)
• Image recognition(image-cat/dog/mug/hat… labled)
• unsupervised : un-labeled data
• word clustering
• google news grouping

 

Supervised learning

• Image labeling : tagged images
• Email spam filter : labeled spam/email
• predicting exam score : previous exam score and time spent

 

Training data set

• {X(feature), Y(label)} 학습하여 모델 생성
• AlphaGo
• 바둑판 기보 학습

Type of supervised learning

• 정확한 시험 점수 예측 0~100점 : regression(넓은 범위)
• pass/fail : 분류, binary classification
• 등급 책정 : multi-label classification