일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | ||||||
2 | 3 | 4 | 5 | 6 | 7 | 8 |
9 | 10 | 11 | 12 | 13 | 14 | 15 |
16 | 17 | 18 | 19 | 20 | 21 | 22 |
23 | 24 | 25 | 26 | 27 | 28 |
Tags
- ChatGPT 설명
- Discrete Wavelet Transform
- binary search
- 이진탐색
- haar matrix
- 선택정렬 증명
- GPT-3
- Binary Search Proof
- 이진탐색 증명
- Proof Selection Sort
- BERT
- Selection Sort
- 선택정렬
- chatGPT
Archives
- Today
- Total
Just Do IT
ML의 간단한 종류 정리 본문
본 포스팅은 유튜브 이수안컴퓨터연구소의 강의를 정리하여 만들었습니다.
https://www.youtube.com/watch?v=fVUinLbWBeQ&list=PL7ZVZgsnLwEF6iyCXbinHLm9nhokrPjdy&index=11
1. 회귀 알고리즘(Regression Algorithms)
- 모델에 의한 예측을 위해 오차 측정을 사용하여 반복적으로 구체화된 변수 간의 관계를 모델링
- 통계의 핵심이며 통계 기반 기계학습에서 채택
- 알고리즘
- Ordinary Least Squared Regression(OLSR)
- Linear Regression
- Logistic Regression
- Stepwise Regression
- Multivariate Adaptive Regression Splines(MARS)
- Locally Estimated Scatterplot Smoothing(LOESS)
2. 인스턴스 기반 알고리즘(Instance-based Algorithms)
- 모델에 중요하거나 필요하다고 여기는 학습 데이터의 인스턴스 또는 예제에 대한 의사결정
- 예측을 위해 유사성 측정 등을 사용해 비교
- 알고리즘
- k-Nearest Neighbor(KNN)
- Learning Vector Quantization(LVQ)
- Self-Organizing Map (SOM)
- Locally Weighted Learning (LWL)
- Support Vector Machines(SVM)
3. 정규화 알고리즘 (Regularization Algorithms)
- 복잡성에 따라 모델에 불이익을 주는 방법으로 일반화
- 일반적으로 회귀 방법 등에 확장되어 사용
- 알고리즘
- Ridge Regression
- Least Absolute Shrinkage and Selection Operator(LASSO)
- Elastic Net
- Least-Angle Regression(LARS)
3. 의사 결정 트리 알고리즘(Decision Tree Algorithms)
- 데이터 속성의 실제 값을 기반으로 의사 결정 모델 구성
- 주어진 레코드에 대한 예측을 트리 구조에서 결정
- 분류 및 회귀 문제에 대한 데이터를 학습
- 알고리즘
- Classification and Regression Tree(CART)
- Iterative Dichotomies3(ID3)
- C4.5 and C5.0 (different versions of a powerful approach)
- Chi-squared Automatic Interaction Detection (CHAID)
- Decision Stump
- M5
- Conditional Decision Trees
4. 베이지안 알고리즘(Bayesian Algorithms)
- 베이지안 확률 기반으로 불확실성에 대해 새로운 사건을 추정
- 분류 및 회귀와 같은 문제에 자주 사용
- 알고리즘
- Naive Bayes
- Gaussian Naive Bayes
- Multinomial Naive Bayes
- Averaged One-Dependence Estimators(AODE)
- Bayesian Belief Netword (BBN)
- Bayesian Netword(BN)
5. 클러스터링 알고리즘(Clustering Algorithms)
- Centroid 기반 및 계층 기반의 모델 방법으로 구성
- 데이터를 최대 공통 그룹으로 잘 구성되도록 동작
- 알고리즘
- k-Means
- k-Medians
- Expectation Maximization(EM)
- Hierarchical Clustering
6. 연관 규칙 알고리즘(Association Rule Learning Algorithms)
- 변수 간 관찰된 관계를 가장 잘 설명하는 규칙을 추출
- 어떤 항목이 어떤 항목을 동반하여 등장하는 지를 파악
- 중요하고 유용한 연관성을 발견
- 알고리즘
- Apriori algorithm
- Eclat algorithm
7. 신경망 알고리즘(Artificial Neural Netword Algorithms)
- 생물학적 신경망의 구조 또는 기능에서 영감을 받은 모델
- 회귀와 분류 문제에 일반적으로 사용되는 패턴 매칭의 한 종류
- 실제로는 모든 유형의 문제에 대한 다양한 알고리즘과 변화로 구성된 영역
- 알고리즘
- Perceptron
- Multilayer Perceptrons(MLP)
- Back-Propagation
- Stochastic Gradient Descent
- Hopfield Network
- Radial Basis Function Network(RBFN)
8. 딥러닝 알고리즘(Deep Learning Algorithms)
- 인공신경망의 발전된 형태로 값싼 대규모 연산을 이용하는 방법
- 훨씬 크고 더 복잡한 신경망 구조를 가지고 이미지, 텍스트, 오디오, 비디오와 같은 매우 큰 데이터와 관련
- 알고리즘
- Convolutional Neural Network(CNN)
- Recurrent Neural Networks(RNN)
- Long Short-Term Memory Networks(LSTM)
- Stacked Auto-Encoders
- Deep Boltzmann Machine(DBM)
- Deep Belief Networks(DBN)
9. 차원 감소 알고리즘(Dimensional Reduction Algorithms)
- 데이터의 고유한 구조를 이용하여 데이터를 요약하거나 기술하는 방법
- 고차원의 데이터를 중요한 요소를 포함한 저차원으로 변환
- 알고리즘
- Principal Component Analysis(PCA)
- Principal Component Regression(PCR)
- Partial Least Squares Regression(PLSR)
- Multidimensional Scaling(MDS)
- Linear Discriminant Analysis(LDA)
- Mixture Discriminant Analysis(MDA)
- Quadratic Discriminant Analysis(QDA)
- Flexible Discriminant Analysis(FDA)
10. 앙상블 알고리즘(Ensemble Algorithms)
- 독립적으로 훈련되고 전체적인 예측을 위해 어떤 식으로든 예측이 결합되는 복수의 약한 모델로 구성된 모델
- 어떤 유형의 약한 학습 모델을 결합시키고 어떤 방식으로 결합할지가 중요
- 매우 강력하고 인기있는 기술
- 알고리즘
- Bootstrapped Aggregation(Bagging)
- AdaBoost
- Weighted Average(Blending)
- Stacked Generalization(Stacking)
- Gradient Boosting Machines(GBM)
- Gradient Boosted Regression Trees(GBRT)
- Random Forest
'AI Study > ML 개념' 카테고리의 다른 글
내가 보려고 만든 Normal equation부터 AdamW까지 optimizer 총정리 (1) | 2023.12.11 |
---|---|
일반화(Generalization), 과대적합(Overfitting), 과소적합(Underfitting) (1) | 2022.02.04 |
이진 분류기의 성능 평가 척도 (Confusion Matrix) (1) | 2022.02.04 |
ML의 간단한 개념 정리 (0) | 2022.02.03 |
Comments