Classification Decision Trees RuleBased Classifiers Nearest Neighbor Classifiers


강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 2 Data Mining & Practices by Yang-Sae Moon

규칙기반 분류기 (Rule-Based Classifier) Page 3 Classification Data Mining & Practices by Yang-Sae Moon

규칙기반 분류기 예제 Classification Reptiles: 파충류 Amphibians: 양서류 Page 4 Data Mining & Practices by Yang-Sae Moon

규칙기반 분류기의 적용 Classification Page 5 Data Mining & Practices by Yang-Sae Moon

규칙 적용범위(Coverage)와 정확도(Accuracy) Page 6 Classification Data Mining & Practices by Yang-Sae Moon

규칙기반 분류기의 동작 방법 Page 7 Classification Data Mining & Practices by Yang-Sae Moon


의사결정 트리 규칙 생성 Page 9 Classification Data Mining & Practices by Yang-Sae Moon

(생성된) 규칙의 단순화 Classification Page 10 Data Mining & Practices by Yang-Sae Moon

규칙 단순화에 의한 효과 영향 Page 11 Classification Data Mining & Practices by Yang-Sae Moon

순서화된 규칙 집합 (Ordered Rule Set) Page 12 Classification Data Mining & Practices by Yang-Sae Moon

분류 규칙의 생성 방법 Classification Page 13 Data Mining & Practices by Yang-Sae Moon

직접 방법: 순차적 커버링(Sequential Covering) Page 14 Classification Data Mining & Practices by Yang-Sae Moon

순차적 커버링 예제 Classification Page 15 Data Mining & Practices by Yang-Sae Moon

순차적 커버링 진행 Classification Rule Growing Instance Elimination Rule Evaluation Stopping Criterion Rule Pruning Page 16 Data Mining & Practices by Yang-Sae Moon

순차적 커버링 - Rule Growing Classification hibernate: 동면 Page 17 Data Mining & Practices by Yang-Sae Moon

순차적 커버링 - Instance Elimination Classification Why do we need to eliminate instances? Otherwise, the next rule is identical to previous rule Page 18 Data Mining & Practices by Yang-Sae Moon

순차적 커버링 - Rule Evaluation Classification Metrics – Accuracy n : Number of instances nc : Number of instances covered by rule – Laplace k : Number of classes p : Prior probability – M-estimate Page 19 Data Mining & Practices by Yang-Sae Moon

순차적 커버링 - Stopping Criterion & Rule Pruning Page 20 Classification Data Mining & Practices by Yang-Sae Moon

간접 방법: 의사결정 트리 등 사용 Page 21 Classification Data Mining & Practices by Yang-Sae Moon

규칙기반 분류기의 장점 Classification Page 22 Data Mining & Practices by Yang-Sae Moon

강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 23 Data Mining & Practices by Yang-Sae Moon

인스턴스 기반 분류기 (1/2) Page 24 Classification Data Mining & Practices by Yang-Sae Moon

인스턴스 기반 분류기 (2/2) Classification rote: (기계적) 암기 Page 25 Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 (Nearest Neighbor Classifiers) Page 26 Classification Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 개념 Classification Page 27 Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 정의 Classification Page 28 Data Mining & Practices by Yang-Sae Moon

1 인접 이웃 (1 -Nearest Neighbor) Page 29 Classification Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 이슈 (1/4) Page 30 Classification Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 이슈 (2/4) Page 31 Classification Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 이슈 (3/4) Page 32 Classification Data Mining & Practices by Yang-Sae Moon

인접 이웃 분류기 이슈 (4/4) Page 33 Classification Data Mining & Practices by Yang-Sae Moon

강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 34 Data Mining & Practices by Yang-Sae Moon

베이지안 분류기 (Bayesian Classifiers) Page 35 Classification Data Mining & Practices by Yang-Sae Moon

베이스 정리의 예제 Classification meningitis: 뇌막염 stiff neck: 뻣뻣한 목 Page 36 Data Mining & Practices by Yang-Sae Moon

베이지안 분류기 개념 (1/2) Page 37 Classification Data Mining & Practices by Yang-Sae Moon

베이지안 분류기 개념 (2/2) Page 38 Classification Data Mining & Practices by Yang-Sae Moon

순수 베이지안 분류기 (Naïve Bayes Classifier) Page 39 Classification Data Mining & Practices by Yang-Sae Moon

훈련 집합에서 확률 구하기 (1/3) Page 40 Classification Data Mining & Practices by Yang-Sae Moon

훈련 집합에서 확률 구하기 (2/3) Page 41 Classification Data Mining & Practices by Yang-Sae Moon

훈련 집합에서 확률 구하기 (3/3) Page 42 Classification Data Mining & Practices by Yang-Sae Moon

순수 베이지안 분류기 예제 (1/2) Page 43 Classification Data Mining & Practices by Yang-Sae Moon

순수 베이지안 분류기 예제 (2/2) Page 44 Classification Data Mining & Practices by Yang-Sae Moon

베이지안 분류기 요약 Classification Page 45 Data Mining & Practices by Yang-Sae Moon

강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 46 Data Mining & Practices by Yang-Sae Moon

Thinking Machine from Brain Page 47 Classification Data Mining & Practices by Yang-Sae Moon

Activation Function – Neuron? Cell? Page 48 Classification Data Mining & Practices by Yang-Sae Moon

인공 신경망 개념 (1/3) Classification Page 49 Data Mining & Practices by Yang-Sae Moon

인공 신경망 개념 (2/3) Classification Page 50 Data Mining & Practices by Yang-Sae Moon

인공 신경망 개념 (3/3) Classification Page 51 Data Mining & Practices by Yang-Sae Moon

인공 신경망의 일반적 구조 Page 52 Classification Data Mining & Practices by Yang-Sae Moon

인공 신경망의 학습 알고리즘 Page 53 Classification Data Mining & Practices by Yang-Sae Moon

인공신경망의 하드웨어 구현 Page 54 Classification Data Mining & Practices by Yang-Sae Moon

더 많은 레이어 딥 러닝 Page 55 Classification Data Mining & Practices by Yang-Sae Moon

Convolutional NN – CNN (1/2) Page 56 Classification Data Mining & Practices by Yang-Sae Moon

Convolutional NN – CNN (2/2) Page 57 Classification Data Mining & Practices by Yang-Sae Moon

순차적 성질? Recurrent NN Page 58 Classification Data Mining & Practices by Yang-Sae Moon

머신러닝 강좌 Classification HKUST 김성훈 교수님 https: //hunkim. github. io/ml/ NN 관련 많은 슬라이드가 위 강의 사이트에서 발췌되었음 Page 59 Data Mining & Practices by Yang-Sae Moon

강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 60 Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (1/7) Page 61 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (2/7) Page 62 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (3/7) Page 63 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (4/7) Page 64 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (5/7) Page 65 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (6/7) Page 66 Classification Data Mining & Practices by Yang-Sae Moon

SVM (Support Vector Machines) (7/7) Page 67 Classification Data Mining & Practices by Yang-Sae Moon

비선형 SVM (Nonlinear SVM) (1/2) Page 68 Classification Data Mining & Practices by Yang-Sae Moon

비선형 SVM (Nonlinear SVM) (2/2) Page 69 Classification Data Mining & Practices by Yang-Sae Moon

강의 내용 Classification 분류 정의와 적용사례 의사결정 트리 (Decision Trees) 규칙기반 분류기 (Rule-Based Classifiers) 인접 이웃 분류기 (Nearest Neighbor Classifiers) 베이지안 분류기 (Bayesian Classifiers) 인공 신경망 (Artificial Neural Networks) 지지도 벡터 머신 (Support Vector Machines) Page 70 Data Mining & Practices by Yang-Sae Moon
- Slides: 70