# Machine-Learning-Toolbox

**Repository Path**: xxfeizai/Machine-Learning-Toolbox

## Basic Information

- **Project Name**: Machine-Learning-Toolbox
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-06-28
- **Last Updated**: 2021-06-28

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Machine-Learning-Toolbox

## 1.DataToolBox
### (1) Metric:
acc=get_acc(real_label,predict_label)

auc=get_auc(real_label, scores)

[tpr,tnr,macc]=get_macc(real_label, predict_label)

### (2) Re-sampling:
[data_much,data_less,much_label,less_label]=divide_data(data, label)

Random Under-Sampling:
[train_data_temp,train_label_temp]=RUS(data_much, data_less, much_label, less_label)

Random Over-Sampling:
[train_data_temp,train_label_temp]=ROS(data_much, data_less, much_label, less_label)

## 2.Ensemble
### (1) Stacking
stacking(clfs,X_train,y,X_test,nfolds=5,stage=1,random_seed=2017,shuffle=True,clfs_name=None,final_clf=None)

## 3. MatMHKS
A matrix based linear classifier

clf=MatMHKS(penalty='l2', C=1.0, matrix_type=None,class_weight=None, max_iter=100,u0=0.5,b0=10**(-6),eta=0.99,min_step=0.0001,multi_class='ovr', verbose=0)

clf.fit(X,y)

clf.predict(X)

clf.predict_proba(X)

## 4.generator
generator for keras/tensorflow (enhanced,imbalanced data)

## 5.metrics
tpr tnr for keras/tensorflow

## 6.loss

### (1) focal loss

### (2) center loss

### (3) triplet loss

### (4) island loss