• 3회차 (1/31) :

  1. 3회차 (1/31) : Model Evaluation(classification) 담당 : 유응규 나지혜
     • Evaluation Methods
       • Hold-out
       • Cross Validation 담당 : 이지인 신광욱
     • Evaluation Metrics(confusion matrix, ROC, AUC)
     • Evaluation Metrics(분산, MAE, R2)

REF

• 각자 자료를 조사하면서 출처 혹은 참고하면 좋을만한 자료를 여기에 남겨주세요

• 유투브 주소)

Backgrounds

Model Parameters and Hyperparameters

Overfitting & Underfitting

Variance &Bias

Evalueation Methods

Hold-out(Training and Test)

Hold - out (Trainging, Validatian and Test)

• Workflow
• disadvantages

Cross Validation

K-fold Cross Validation

---

Model Tuning

Learning Curves

Validation Curves

Hyperparameter Optimization

Gridsearch

Randomized Search

Nested Cross validation

Evaluation Metrics

• Classification

  Confusin Matrics

  ROC

  AUC

• Regression

  EV

  MAE

  MSE

  MSLE

  R squared