Section 2 종합 회고

SECTION 2 리뷰
01 - Linear Models
02 - Tree Based Model
03 - Applied Predictive Modeling
04 - Project (NBA 게임 결과 예측 머신러닝 모델)

01 - KEYWORD

• 선형회귀 (Linear Regression)
• 지도학습 (Supervised Learning)
• 기준모델 (Baseline Model) (분류문제-타겟의 최빈 클래스) (회귀문제-평균값) (시계열문제-이전 타임스탬프의값)
• 회귀선
• 잔차
• RSS (Residual sum of squares) (Sum of Square Error)
• 비용함수
• 최소제곱회귀 (OLS , Ordinary least squares)
• 종속변수 (반응변수 , 레이블 , 타겟)
• 독립변수 (예측변수 , 설명 , 특성)
• 단순 선형 회귀 (Simple Linear Regression)
• 계수 (Coefficients)
• 절편 (Intercept)
• 다중 선형 회귀 (Multiple Linear Regression)
• MSE (Mean Squared Error)
• MAE (평균절대오차 , Mean Absolute Error)
• RMSE (Root Mean Squared Error)
• R-squared (Coefficient of determination)
• 일반화
• 과적합
• 과소적합
• 편향
• 분산
• 트레이드오프
• Ridge Regression
• Ridge CV - alpha , lambda
• One hot encoding
• 특성선택 (Feature Selection) _ 특성공학 , 과제에 적합한 특성을 만들어 내는 과정
• 정규화 (regularization)
• 카테고리형 데이터 - 명목형(nominal) , 순서형(ordinal)
• 분류문제
• Logistic Regression
• 훈련 / 검증 / 테스트 세트
• 하이퍼파라미터
• 교차검증 (K-fold corss validation)
• 오즈(Odds . 실패확률에 대한 성공확률의 비)
• 로짓 변환 (Logit transformation)

Linear Regression Analysis,선형모델 첫시간_Day21	khalidpark2029.tistory.com/67
Tabular Data, 분류와 회귀_Day21(2)	khalidpark2029.tistory.com/68
단순선형회귀,Simple Linear Regression_Day21(3)	khalidpark2029.tistory.com/69
R Squared 계산방법 , R 스퀘어, 결정계수_Day22	khalidpark2029.tistory.com/70
Mean Square Error,평균 제곱근 편차,잔차와오차_Day22(2)	khalidpark2029.tistory.com/71
Training & Test , Bias & Variance,편향,분산_Day22(3)	khalidpark2029.tistory.com/72
과적합(Overfitting)과 과소적합(Underfitting)_Day22(4)	khalidpark2029.tistory.com/73
Bias/Variance/편향과분산, 한번더_Day22(5)	khalidpark2029.tistory.com/75
회귀의 오류지표,MAE,MSE,RMSE,MAPE,MPE_Day22(6)	khalidpark2029.tistory.com/76
Regularization,Ridge,릿지,regression_Day23	khalidpark2029.tistory.com/77
다중 회귀 분석 vs 다항 회귀 분석_Day23(2)	khalidpark2029.tistory.com/78
fit_transform과 transform의 차이,싸이킷런	khalidpark2029.tistory.com/82
회귀의 종류 총정리	khalidpark2029.tistory.com/83
Regularization,Lasso ,라쏘,regression_Day23(3)	khalidpark2029.tistory.com/85
Regularization,Elastic Net,엘라스틱넷,regression_Day23(4)	khalidpark2029.tistory.com/86
Cross Validation(CV)와 lambda , Regression_Day23(5)	khalidpark2029.tistory.com/87
Overfitting, train, validation and test_Day24	khalidpark2029.tistory.com/88
Logistic Regression,로지스틱_Day24(2)	khalidpark2029.tistory.com/89
Logistic Regression ,Coefficients_Day24(3)	khalidpark2029.tistory.com/90
Logistic Regression,Maximum Likelihood_Day24(4)	khalidpark2029.tistory.com/91

 

02 - KEYWORD

• 결정트리 (Decision Tree)
• 노드(node)
• 뿌리노드 (root) , 중간노드 (internal) , 말단노드 (external, leaf, terminal)
• 엣지(edge)
• 지니 불순도 (gini impurity)
• 파이프라인(Pipelines)
• 결정트리 과적합 문제 해결 방법 (min_samples_split , min_samples_leaf , max_depth)
• 특성 중요도(feature importances)
• 랜덤포레스트(random forest)
• 배깅(bagging)
• 부트스트랩(bootstrap)
• Aggregation - 부트스트랩세트로 만들어진 기본모델들을 합치는 과정
• Ordinal encoding
• 혼동행렬 (Confusion matrix)
• 정확도
• 정밀도
• 재현율
• 임계값
• ROC curve
• AUC (ROC 면적)
• 모델선택(Model Selection)
• 교차검증(Cross Validation)
• 하이퍼파라미터 튜닝 _최적화 , 일반화
• 검증곡선 (Validation curve)
• RandomizedSearchCV
• GridSearchCV
  
  

결정트리,Decision Trees_Day26	khalidpark2029.tistory.com/92
Random Forest,랜덤포레스트_Day27	khalidpark2029.tistory.com/94
Confusion Matrix, 혼동행렬_Day28	khalidpark2029.tistory.com/101
결정트리,랜덤포레스트,혼동행렬,교차검증 키워드 개념정리	khalidpark2029.tistory.com/104
혼동행렬, 정확도, 정밀도, 재현율 한번더정리	khalidpark2029.tistory.com/105

 

03 - KEYWORD

• 정보누수 (leakage)
• Data wrangling
• 특성중요도 (feature importances)
• 1) MDI (Mean decrease impurity)
• 2) Drop Column Importance
• 3) 순열중요도 (MDA , Mean decrease accuracy)
• eli5 라이브러리
• Adaboost
• XGboost
• LightGBM
• PDP (부분의존도그림 , Partial dependence plot)
• SHAP

분류 정확도와 불균형, accuracy만 집중했을 때의 문제점_Day 31	khalidpark2029.tistory.com/107
판다스 groupby 활용법_Day 32	khalidpark2029.tistory.com/108
bootstrap,bagging 복습_Day33	khalidpark2029.tistory.com/109
AdaBoost,아다부스트,decision tree,random tree_Day33(2)	khalidpark2029.tistory.com/110
Gradient Boost,for Regression_Day33(3)	khalidpark2029.tistory.com/113
부분의존도,Partial Dependence Plot (PDP)_Day34	khalidpark2029.tistory.com/114

 

04 - Project

• NBA 머신러닝 예측 모델 프로젝트

 

 

프로젝트)NBA 머신러닝 예측 모델_Day36~39	khalidpark2029.tistory.com/115
완성)NBA 머신러닝 예측 모델 프로젝트_Day40	khalidpark2029.tistory.com/119
피드백)NBA 머신러닝 예측 모델 프로젝트_Day40	khalidpark2029.tistory.com/120