從零基礎入門Tensorflow2.0 ----5.3 實戰sklearn超參數搜索

every blog every motto: Until you make peace with who you are, you’ll never be content with what you have.

0. 前言

實戰sklearn超參數搜索。
注： 訓練時間較長。

1. 代碼部分

1. 導入模塊

import matplotlib as mpl
import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
import sklearn
import pandas as pd
import os
import sys
import time
import tensorflow as tf
from tensorflow import keras

print(tf.__version__)
print(sys.version_info)
for module in mpl,np,pd,sklearn,tf,keras:
    print(module.__name__,module.__version__)

2. 讀取數據

from sklearn.datasets import fetch_california_housing

# 房價預測
housing = fetch_california_housing()
print(housing.DESCR)
print(housing.data.shape)
print(housing.target.shape)

3. 劃分樣本

# 劃分樣本
from sklearn.model_selection import train_test_split

x_train_all,x_test,y_train_all,y_test = train_test_split(housing.data,housing.target,random_state=7)
x_train,x_valid,y_train,y_valid = train_test_split(x_train_all,y_train_all,random_state=11)

print(x_train.shape,y_train.shape)
print(x_valid.shape,y_valid.shape)
print(x_test.shape,y_test.shape)

4. 數據歸一化

# 歸一化
from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
x_train_scaled = scaler.fit_transform(x_train)
x_valid_scaled = scaler.transform(x_valid)
x_test_scaled = scaler.transform(x_test)

5. 構建模型、訓練

RandomizedSearchCV
步驟：

1. 轉換爲sklearn的model（上節實現）
2. 定義參數集合（本節實現）
3. 搜索參數（本節實現）

# RandomizedSearchCV
# 步驟
# 1. 轉換爲sklearn的model
# 2. 定義參數集合
# 3. 搜索參數

def build_model(hidden_layers=1,layer_size=30,learning_rate=3e-3):
    model = keras.models.Sequential()
    model.add(keras.layers.Dense(layer_size,activation='relu',input_shape=x_train.shape[1:]))
    
    for _ in range(hidden_layers - 1):
        model.add(keras.layers.Dense(layer_size,activation='relu'))
        
    model.add(keras.layers.Dense(1))
    optimizer = keras.optimizers.SGD(learning_rate)
    model.compile(loss="mse",optimizer=optimizer)
        
    return model

# 轉成sklearn model
sklearn_model = keras.wrappers.scikit_learn.KerasRegressor(build_model)

# 回調函數
callbacks = [keras.callbacks.EarlyStopping(patience=5,min_delta=1e-3)]
# 訓練
history = sklearn_model.fit(x_train_scaled,y_train,epochs=100,validation_data=(x_valid_scaled,y_valid),callbacks=callbacks)

6. 學習曲線

# 學習曲線
def plot_learning_curves(history):
    pd.DataFrame(history.history).plot(figsize=(8,5))
    plt.grid(True)
    plt.gca().set_ylim(0,1)
    plt.show()
plot_learning_curves(history)

7. 超參數搜索

cross_validation:訓練集分成n份，n-1訓練，最後一份驗證

from scipy.stats import reciprocal
# f(x) = 1/(x*log(b/a)) a<=x <=b
param_distribution = {
    "hidden_layers":[1,2,3,4],
    "layer_size":np.arange(1,100),
    "learning_rate":reciprocal(1e-4,1e-2),
}

from sklearn.model_selection import RandomizedSearchCV

random_search_cv = RandomizedSearchCV(sklearn_model,param_distribution,n_iter = 10,n_jobs=1)
random_search_cv.fit(x_train_scaled,y_train,epochs=100,validation_data=(x_valid_scaled,y_valid),callbacks=callbacks)
# cross_validation:訓練集分成n份，n-1訓練，最後一份驗證

8. 查詢最好參數、分值、模型

# 查詢最好參數、分值、模型
print(random_search_cv.best_params_)
print(random_search_cv.best_score_)
print(random_search_cv.best_estimator_)

9. 獲取最好模型，並在測試集上驗證

# 獲取最好模型，對測試集進行測試
model = random_search_cv.best_estimator_.model
model.evaluate(x_test_scaled,y_test)