机器学习算法源码实现。
Contributer: datamonday
Github Repo: https://github.com/datamonday/ML-Algorithm-Source-Code
-
- kNN (k-Nearest-Neighbors)
-
- Logistic Regression
-
- Gaussian Discriminat Analysis
-
- Naive Bayes
-
Decision Tree
- ID3
- C4.5
- CART (Classification and Regression Decision Tree)
-
SVM (Support Vector Machine)
- SVC (Support Vector Classifier)
- SVR (Support Vector Regression)
-
Linear Regression
- PolynomialRegression
- LassoRegression (L1)
- RidgeRegression (L2)
- ElasticNet (L1+L2)
-
Bagging
- Random Forest
-
Boosting
- AdaBoost
- GBDT
- XGBoost
-
Stacking
-
Perceptron
- Simple Perceptron
- Multi-Layer Perceptron
-
LDA (Linear Discriminant Analysis)
- Simple LDA
- Multi-Class LDA
-
- PCA (Principal Component Analysis)
- K-Means
- K-Means
- K-Means ++
- ISODATA
-
- Fuzzy C-Means
-
- Gaussian Mixed Model
-
- SOFM (Self-Organized Feature Map)
-
- DBSCAN
- AutoEncoder (AE)
- AE
- VAE
- GAN (Generative Adversarial Network)
- DCGAN
- StyleGAN
- Restricted Boltzmann Machine (RBM)
- Anormaly Detection
- Isolation Forest
- One-Class SVM
-
- GA (Genetic Algorithm)
-
- PSO (Particle Swarm Optimization)
-
- ACO (Ant Clony Optimization)
-
- SA (Simulated Annealing)
-
- NeuroEvolution
-
Activation Functions
- Sigmoid
- Tanh
- Softmax
- ReLU
- LeakyReLU
- PReLU
- ELU
- SELU
- SoftPlus
-
Loss Functions
- HigeLoss
- SquareLoss
- CrossEntropy
- MSE
- RMSE
- RMSLE
- MAE
- MAPE
-
Optimizers
- GD (Gradient Descent)
- SGD (Stochastic Gradient Descent)
- Mini-GD
- NAG (Nesterov Accelerated Gradient)
- SGD + Momentum
- AdaGrad
- AdaDelta
- RMSProp
- Adam
- AdaMax
- Nadam (Adam + NAG)
-
Layers
- Dense
- Flatten
- Reshape
- Dropout
- Activation
- Batch Normalization
- Layer Normalization
- Group Normalization
- RNN
- LSTM
- Conv1D
- Conv2D
- MaxPooling2D
- AvgPooling2D
-
Neural Networks
- Q-Learning
- Sara-Learning
-
Kernels
- Linear Kernel
- Polynomial Kernel
- RBF Kernel
-
Data Pipeline
-
Online Flow Data Process
-
Data Operation
- shuffle data
- normalize
- standardize
- batch iterator
- divide on features
- polynomial features
- get random subsets
- normalize
- standardize
- train test split
- k fold cross validation sets
- bootstrap sample
- to categorical
- to nominal
- make diagonal
-
Evaluation Metrics
- calculate_entropy
- mean_squared_error
- calculate_std_dev
- calculate_variance
- accuracy_score
- recall_score
- precision_score
- f1_score
- calculate_covariance_matrix
- calculate_correlation_matrix
-
Feature Selection
- Filter
- wrapper
- statistical
-
Fine-Tune Hyperparameters
- Grid Search
- Random Search
- Bayesian Optimization
- Hyperband
-
PlotFunctions
- loss curve
- residual loss curve
- acc curve
- roc
- p-r curve