kaggle1 - 决策树

 

# 数据预处理

　data.dropna(axis=0)　　# 取出掉缺失值 按行

melbourne_features = ['Rooms', 'Bathroom', 'Landsize', 'Lattitude', 'Longtitude']

X = melbourne_data[melbourne_features]

X.describe()# 描述一下数据

# 分隔数据

from sklearn.model_selection import train_test_split
train_X, val_X, train_y, val_y = train_test_split(X, y, random_state = 0)　　# 随机数有利于重现结果

# 训练数据

　　from sklearn.model_selection import train_test_split
　　# Specify the model
　　iowa_model = DecisionTreeRegressor(random_state=1)

　　# Fit iowa_model with the training data.
　　iowa_model.fit(train_X,train_y)

# 对预测集数据作出预测

　　# Predict with all validation observations
　　val_predictions = iowa_model.predict(val_X)

# 计算模型的误差

　　from sklearn.metrics import mean_absolute_error
　　val_mae = mean_absolute_error(val_predictions,val_y)

　　# uncomment following line to see the validation_mae
　　print(val_mae)

# 定义一个方法去,尝试不同的参数,来优化你的模型

def get_mae(max_leaf_nodes, train_X, val_X, train_y, val_y):

model = DecisionTreeRegressor(max_leaf_nodes=max_leaf_nodes, random_state=0)
model.fit(train_X, train_y)
preds_val = model.predict(val_X)
mae = mean_absolute_error(val_y, preds_val)
return(mae)

# 最后决定好了,你就不需要,数据来做验证了,全部用来训练吧! 

 

 

# 建立比决策树更好的模型,随机森林 查看一下误差
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_absolute_error

forest_model = RandomForestRegressor(random_state=1)
forest_model.fit(train_X, train_y)
melb_preds = forest_model.predict(val_X)
print(mean_absolute_error(val_y, melb_preds))