Catboost介绍

1.简介

Catboost是基于梯度提升的决策树模型。

优点：

• 和GBDT库比，具有更好的性能；
• 最好的分类预测速度；
• 同时支持数值型和类别型的特征；
• 更快的GPU和多GPU支持；
• 包含可视化的工具。

2 分类实例

from catboost import Pool, CatBoostClassifier

import numpy as np 
import matplotlib.pyplot as plt
#生成用于分类的数据集
from sklearn.datasets.samples_generator import make_classification
X,labels=make_classification(n_samples=2000,n_features=20,n_redundant=0,n_informative=4,n_classes= 3, 
							 random_state=1,n_clusters_per_class=2)
rng=np.random.RandomState(2)
X+=2*rng.uniform(size=X.shape)
labels = np.reshape(labels, (len(labels),-1))
X_y = np.concatenate([X,labels], axis = 1)
train_num = int(len(X)*0.8)
train_data = X_y[:train_num,:len(X_y[0])-1]
eval_data = X_y[train_num:, :len(X_y[0])-1]

cat_features = []

train_label = X_y[:train_num, len(X_y[0])-1]
eval_label = X_y[train_num:, len(X_y[0])-1]
print(train_data.shape)
print(train_data)
print(eval_data.shape)
print(train_label.shape)
print(eval_label.shape)

train_dataset = Pool(data=train_data,
                     label=train_label,
                    )

eval_dataset = Pool(data=eval_data,
                    label=eval_label,
                    )

# Initialize CatBoostClassifier
model = CatBoostClassifier(iterations=10,
                           learning_rate=1,
                           depth=2,
                           loss_function='MultiClass')
# Fit model
model.fit(train_dataset)
# Get predicted classes
preds_class = model.predict(eval_dataset)
print(preds_class)
# Get predicted probabilities for each class
preds_proba = model.predict_proba(eval_dataset)
# Get predicted RawFormulaVal
preds_raw = model.predict(eval_dataset, 
                          prediction_type='RawFormulaVal')
print(preds_proba)

---

参考：

1. catboost官网；
2. catboost usage examples;
3. CatBoost: unbiased boosting with categorical features;