唐宇迪课程学习笔记
<>回归问题预测
* Tensordlow2版本中将大量使用keras的简介建模方法 import numpy as np import pandas as pd
import marplotlib.pyplot as plt import tensorflow as tf from tensorflow.keras
import layers import tensorflow.keras import warnings warnings.filterwarnings(
"ignore") %matplotlib inline features = pd.read_csv('temps.csv') # 看看数据长什么样子
features.head()
数据表中
* year,month,day,week 分别表示具体的时间
* temp_2:前天的最高气温
* temp_1:昨天的最高气温
* average:在历史中,每年这一天的平均最高温度值
* actual:这就是我们的标签值了,当天的真实最高温度
* friend:这一列可能是凑热闹的,你的朋友猜测的可能值,咱们不管它就好 print('数据维度:', features.shape)
数据维度:(348, 9)
# 处理时间数据 import datetime # 分别得到年,月,日 years = features['year'] months = features
['month'] days = features['day'] # datetime 格式 dates = [str(int(year)) + '-' +
str(int(month)) + '-' + str(int(day)) for years, month, day in zip(years, months
, days)] dates = [datetime.datetime.strptime(date, '%Y-%m-%d') for date in dates
] dates[:5]
[datatime.datetime(2016, 1, 1, 0, 0),
datatime.datetime(2016, 1, 2, 0, 0),
datatime.datetime(2016, 1, 3, 0, 0),
datatime.datetime(2016, 1, 4, 0, 0),
datatime.datetime(2016, 1, 5, 0, 0)]
# 准备画图 # 指定默认风格 plt.style.use('fivethirtyeight') # 设置布局 fig, ((ax1, ax2), (ax3,
ax4)) = plt.subplots(nrows=2, cols=2, figsize = (10, 10)) fig.autofmt_xdate(
rotation= 45) # 标签值 ax1.plot(dates, features['actual']) ax1.set_xlabel('');ax1.
set_ylabel('Temperature');ax1.set_title('MAX Temp') # 昨天 ax2.plot(dates,
features['tenp_1']) ax2.set_xlabel('');ax2.set_ylabel('Temperature');ax2.
set_title('Previous Max Temp') # 前天 ax3.plot(dates, features['temp_2']) ax3.
set_xlabel('Date');ax3.set_ylabel('Temperature');ax3.set_title('Two Days Prior
Max Temp') # 我的逗逼朋友 ax4.plot(dates, features['friend']) ax4.set_xlabel('Date');
ax4.set_ylabel('Temperature');ax4,set_title('Friend Estimate') plt.tight_layout(
pad=2)
# 独热编码 features = pd.get_dumies(features) features.head(5)
labels = np.array(features['actual']) # 在特征中去掉标签 features = features.drop(
'actual', axis=1) # 名字单独保存一下,以备后患 feature_list = list(features.columns) #
转换成合适的格式 features = np.array(features) features,shape
(348, 14)
from sklearn import preprocessing input_features = preprocessing.StandardScaler
().fit_transform(features)
<>基于Keras构建网络模型
一些常用参数已经列出,如下所示:
* activation:激活函数的选择,一般常用relu
* kernel_initializer, bias_initializer:权重与偏置参数的初始化方法,有时候不收敛换种初始化突然好使了…玄学
* kernel_regularizer, bias_regularizer:要不要加入正则化
* inputs:输入,可以自己制定,也可以让网络自动选
* units:神经元个数
按顺序构造网络模型
model = tf.keras.Sequential() model.add(layers.Dense(16)) model.add(layers.
Dense(32)) model.add(layers.Dense(1))
compile 相当于对网络进行配置,指定好优化器和损失函数等
# SGD 梯度下降 model.compile(optimizer=tf.keras.optimizers.SGD(0.001), loss=
'mean_squared_error') model.fit(input_features, labels, validation_split=0.25,
epochs=10, batch_size=64) # x, y, 验证集占25%, 10轮, 优化器每次迭代64
似乎存在一些问题,模型还没有完全收敛,能不能调些参数呢
model.summary()
<>更改初始化方法后
model = tf.keras.Sequential() model.add(layers.Dense(16, kernel_initializer=
'random_normal')) model.add(layers.Dense(32, kernel_initializer='random_normal')
) model.add(layers.Dense(1, kernel_initializer='random_normal')) model.compile(
optimizer=tf.keras.optimizers.SGD(0.001), loss='mean_squared_error') model.fit(
input_features, labels, validation_split=0.25, epochs=100, batch_size=64)
<>加入正则化惩罚项
model = tf.keras.Sequential() model.add(layers.Dense(16, kernel_initializer=
'random_normal', kernel_regularizer=tf.keras.regularizers.l2(0.03))) model.add(
layers.Dense(32, kernel_initializer='random_normal', kernel_regularizer=tf.keras
.regularizers.l2(0.03))) model.add(layers.Dense(1, kernel_initializer=
'random_normal', kernel_regularizer=tf.keras.regularizers.l2(0.03))) model.
compile(optimizer=tf.keras.optimizers.SGD(0.001), loss='mean_squared_error')
model.fit(input_features, labels, validation_split=0.25, epochs=100, batch_size=
64)
加入正则项,可以使W更加平滑
<>预测模型结果
predict = model.predict(input_features) predict.shape
(348, 1)
<>测试结果并进行展示
# 转换日期 dates = [str(int(year)) + '-' + str(int(month)) + '-' + str(int(day))
for years, month, day in zip(years, months, days)] dates = [datetime.datetime.
strptime(date, '%Y-%m-%d') for date in dates] # 创建一个表格来存日期和其对应的标签数值 true_data =
pd.DataFrame(data = {'date':dates, 'actual':labels}) # 同理,再创建一个来存日期和其对应的模型预测值
months= features[:, feature_list.index('month')] days = features[:, feature_list
.index('day')] year = features[:, feature_list.index('year')] test_dates = [str(
int(year)) + '-' + str(int(month)) + '-' + str(int(day)) for years, month, day
in zip(years, months, days)] tset_dates = [datetime.datetime.strptime(date,
'%Y-%m-%d') for date in dates] predications_data = pd.DataFrame(data = {'date':
dates, 'actual':labels}) # 真实值 plt.plot(ture_data['date'], true_data['actual'],
'b-', label='actual') # 预测值 plt.plot(predictions_data['data'], predictions_data[
'prediction'], 'ro', label ='prediction') plt.xticks(rotation = '60') plt.legend
() # 图名 plt.xlabel('Date);plt.ylabel('Maximum Temperature (F)');plt.title('
Actualand Predicted Values');
主要看验证集和测试集上的 loss 值