赞
踩
♚
作者:沂水寒城,CSDN博客专家,个人研究方向:机器学习、深度学习、NLP、CV
Blog: http://yishuihancheng.blog.csdn.net
传统的线性模型难以解决多变量或多输入问题,而神经网络如LSTM则擅长于处理多个变量的问题,该特性使其有助于解决时间序列预测问题。在接下来的这篇文章中,你将学会如何利用深度学习库Keras搭建LSTM模型来处理多个变量的时间序列预测问题,你会掌握:
如何将原始数据转化为适合处理时序预测问题的数据格式;
如何准备数据并搭建LSTM来处理时序预测问题;
如何利用模型预测。
一、空气污染预测
在这篇博客中,我们将采用空气质量数据集。数据来源自位于北京的美国大使馆在2010年至2014年共5年间每小时采集的天气及空气污染指数。
数据集包括日期、PM2.5浓度、露点、温度、风向、风速、累积小时雪量和累积小时雨量。原始数据中完整的特征如下:
- 1.No 行数
- 2.year 年
- 3.month 月
- 4.day 日
- 5.hour 小时
- 6.pm2.5 PM2.5浓度
- 7.DEWP 露点
- 8.TEMP 温度
- 9.PRES 大气压
- 10.cbwd 风向
- 11.lws 风速
- 12.ls 累积雪量
- 13.lr 累积雨量
我们可以利用此数据集搭建预测模型,利用前一个或几个小时的天气条件和污染数据预测下一个(当前)时刻的污染程度。可以在UCI Machine Learning Repository下载数据集。
Beijing PM2.5 Data Set
https://archive.ics.uci.edu/ml/datasets/Beijing+PM2.5+Data
二、数据处理
在使用数据之前需要对数据做一些处理,待处理部分数据如下:
- No,year,month,day,hour,pm2.5,DEWP,TEMP,PRES,cbwd,Iws,Is,Ir
- 1,2010,1,1,0,NA,-21,-11,1021,NW,1.79,0,0
- 2,2010,1,1,1,NA,-21,-12,1020,NW,4.92,0,0
- 3,2010,1,1,2,NA,-21,-11,1019,NW,6.71,0,0
- 4,2010,1,1,3,NA,-21,-14,1019,NW,9.84,0,0
- 5,2010,1,1,4,NA,-20,-12,1018,NW,12.97,0,0
粗略的观察数据集会发现最开始的24小时PM2.5值都是NA,因此需要删除这部分数据,对于其他时刻少量的缺省值利用Pandas中的fillna填充;同时需要整合日期数据,使其作为Pandas中索引(index)。
下面的代码完成了以上的处理过程,同时去掉了原始数据中“No”列,并将列命名为更清晰的名字。
- from pandas import read_csv
- from datetime import datetime
- # load data
- def parse(x):
- return datetime.strptime(x, '%Y %m %d %H')
- dataset = read_csv('raw.csv', parse_dates = [['year', 'month', 'day', 'hour']], index_col=0, date_parser=parse)
- dataset.drop('No', axis=1, inplace=True)
- # manually specify column names
- dataset.columns = ['pollution', 'dew', 'temp', 'press', 'wnd_dir', 'wnd_spd', 'snow', 'rain']
- dataset.index.name = 'date'
- # mark all NA values with 0
- dataset['pollution'].fillna(0, inplace=True)
- # drop the first 24 hours
- dataset = dataset[24:]
- # summarize first 5 rows
- print(dataset.head(5))
- # save to file
- dataset.to_csv('pollution.csv')
处理后的数据存储在“pollution.csv”文件中,部分如下:
- pollution dew temp press wnd_dir wnd_spd snow rain
- date
- 2010-01-02 00:00:00 129.0 -16 -4.0 1020.0 SE 1.79 0 0
- 2010-01-02 01:00:00 148.0 -15 -4.0 1020.0 SE 2.68 0 0
- 2010-01-02 02:00:00 159.0 -11 -5.0 1021.0 SE 3.57 0 0
- 2010-01-02 03:00:00 181.0 -7 -5.0 1022.0 SE 5.36 1 0
- 2010-01-02 04:00:00 138.0 -7 -5.0 1022.0 SE 6.25 2 0
现在的数据格式已经更加适合处理,可以简单的对每列进行绘图。下面的代码加载了“pollution.csv”文件,并对除了类别型特性“风速”的每一列数据分别绘图。
- from pandas import read_csv
- from matplotlib import pyplot
- # load dataset
- dataset = read_csv('pollution.csv', header=0, index_col=0)
- values = dataset.values
- # specify columns to plot
- groups = [0, 1, 2, 3, 5, 6, 7]
- i = 1
- # plot each column
- pyplot.figure()
- for group in groups:
- pyplot.subplot(len(groups), 1, i)
- pyplot.plot(values[:, group])
- pyplot.title(dataset.columns[group], y=0.5, loc='right')
- i += 1
- pyplot.show()
运行上述代码,并对7个变量在5年的范围内绘图。
三、多变量LSTM预测模型
3.1 LSTM数据准备
采用LSTM模型时,第一步需要对数据进行适配处理,其中包括将数据集转化为有监督学习问题和归一化变量(包括输入和输出值),使其能够实现通过前一个时刻(t-1)的污染数据和天气条件预测当前时刻(t)的污染。以上的处理方式很直接也比较简单,仅仅只是为了抛砖引玉,其他的处理方式也可以探索,比如:
利用过去24小时的污染数据和天气条件预测当前时刻的污染;
预测下一个时刻(t+1)可能的天气条件;
下面代码中首先加载“pollution.csv”文件,并利用sklearn的预处理模块对类别特征“风向”进行编码,当然也可以对该特征进行one-hot编码。接着对所有的特征进行归一化处理,然后将数据集转化为有监督学习问题,同时将需要预测的当前时刻(t)的天气条件特征移除,完整代码如下:
- # convert series to supervised learning
- def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):
- n_vars = 1 if type(data) is list else data.shape[1]
- df = DataFrame(data)
- cols, names = list(), list()
- # input sequence (t-n, ... t-1)
- for i in range(n_in, 0, -1):
- cols.append(df.shift(i))
- names += [('var%d(t-%d)' % (j+1, i)) for j in range(n_vars)]
- # forecast sequence (t, t+1, ... t+n)
- for i in range(0, n_out):
- cols.append(df.shift(-i))
- if i == 0:
- names += [('var%d(t)' % (j+1)) for j in range(n_vars)]
- else:
- names += [('var%d(t+%d)' % (j+1, i)) for j in range(n_vars)]
- # put it all together
- agg = concat(cols, axis=1)
- agg.columns = names
- # drop rows with NaN values
- if dropnan:
- agg.dropna(inplace=True)
- return agg
-
- # load dataset
- dataset = read_csv('pollution.csv', header=0, index_col=0)
- values = dataset.values
- # integer encode direction
- encoder = LabelEncoder()
- values[:,4] = encoder.fit_transform(values[:,4])
- # ensure all data is float
- values = values.astype('float32')
- # normalize features
- scaler = MinMaxScaler(feature_range=(0, 1))
- scaled = scaler.fit_transform(values)
- # frame as supervised learning
- reframed = series_to_supervised(scaled, 1, 1)
- # drop columns we don't want to predict
- reframed.drop(reframed.columns[[9,10,11,12,13,14,15]], axis=1, inplace=True)
- print(reframed.head())
运行上述代码,能看到被转化后的数据集,数据集包括8个输入变量(输入特征)和1个输出变量(当前时刻t的空气污染值,标签)
数据集的处理比较简单,还有很多的方式可以尝试,一些可以尝试的方向包括:
对“风向”特征哑编码;
加入季节特征;
时间步长超过1。
其中,上述第三种方式对于处理时间序列问题的LSTM可能是最重要的。
3.2 构造模型
在这一节,我们将构造LSTM模型。
首先,我们需要将处理后的数据集划分为训练集和测试集。为了加速模型的训练,我们仅利用第一年数据进行训练,然后利用剩下的4年进行评估。
下面的代码将数据集进行划分,然后将训练集和测试集划分为输入和输出变量,最终将输入(X)
改造为LSTM的输入格式,即[samples,timesteps,features]
。
- # split into train and test sets
- values = reframed.values
- n_train_hours = 365 * 24
- train = values[:n_train_hours, :]
- test = values[n_train_hours:, :]
- # split into input and outputs
- train_X, train_y = train[:, :-1], train[:, -1]
- test_X, test_y = test[:, :-1], test[:, -1]
- # reshape input to be 3D [samples, timesteps, features]
- train_X = train_X.reshape((train_X.shape[0], 1, train_X.shape[1]))
- test_X = test_X.reshape((test_X.shape[0], 1, test_X.shape[1]))
- print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)
运行上述代码打印训练集和测试集的输入输出格式:
(8760, 1, 8) (8760,) (35039, 1, 8) (35039,)
现在可以搭建LSTM模型了。
LSTM模型中,隐藏层有50个神经元,输出层1个神经元(回归问题),输入变量是一个时间步(t-1)
的特征,损失函数采用Mean Absolute Error(MAE),优化算法采用Adam,模型采用50个epochs并且每个batch的大小为72。
最后,在fit()
函数中设置validation_data参数,记录训练集和测试集的损失,并在完成训练和测试后绘制损失图。
- # design network
- model = Sequential()
- model.add(LSTM(50, input_shape=(train_X.shape[1], train_X.shape[2])))
- model.add(Dense(1))
- model.compile(loss='mae', optimizer='adam')
- # fit network
- history = model.fit(train_X, train_y, epochs=50, batch_size=72, validation_data=(test_X, test_y), verbose=2, shuffle=False)
- # plot history
- pyplot.plot(history.history['loss'], label='train')
- pyplot.plot(history.history['val_loss'], label='test')
- pyplot.legend()
- pyplot.show()
-
- # design network
- model = Sequential()
- model.add(LSTM(50, input_shape=(train_X.shape[1], train_X.shape[2])))
- model.add(Dense(1))
- model.compile(loss='mae', optimizer='adam')
- # fit network
- history = model.fit(train_X, train_y, epochs=50, batch_size=72, validation_data=(test_X, test_y), verbose=2, shuffle=False)
- # plot history
- pyplot.plot(history.history['loss'], label='train')
- pyplot.plot(history.history['val_loss'], label='test')
- pyplot.legend()
- pyplot.show()
3.3 模型评估
接下里我们对模型效果进行评估。
值得注意的是:需要将预测结果和部分测试集数据组合然后进行比例反转(invert the scaling),同时也需要将测试集上的预期值也进行比例转换。
通过以上处理之后,再结合RMSE(均方根误差)计算损失。
- # make a prediction
- yhat = model.predict(test_X)
- test_X = test_X.reshape((test_X.shape[0], test_X.shape[2]))
- # invert scaling for forecast
- inv_yhat = concatenate((yhat, test_X[:, 1:]), axis=1)
- inv_yhat = scaler.inverse_transform(inv_yhat)
- inv_yhat = inv_yhat[:,0]
- # invert scaling for actual
- test_y = test_y.reshape((len(test_y), 1))
- inv_y = concatenate((test_y, test_X[:, 1:]), axis=1)
- inv_y = scaler.inverse_transform(inv_y)
- inv_y = inv_y[:,0]
- # calculate RMSE
- rmse = sqrt(mean_squared_error(inv_y, inv_yhat))
- print('Test RMSE: %.3f' % rmse)
整个项目完整代码如下:
- from math import sqrt
- from numpy import concatenate
- from matplotlib import pyplot
- from pandas import read_csv
- from pandas import DataFrame
- from pandas import concat
- from sklearn.preprocessing import MinMaxScaler
- from sklearn.preprocessing import LabelEncoder
- from sklearn.metrics import mean_squared_error
- from keras.models import Sequential
- from keras.layers import Dense
- from keras.layers import LSTM
-
- # convert series to supervised learning
- def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):
- n_vars = 1 if type(data) is list else data.shape[1]
- df = DataFrame(data)
- cols, names = list(), list()
- # input sequence (t-n, ... t-1)
- for i in range(n_in, 0, -1):
- cols.append(df.shift(i))
- names += [('var%d(t-%d)' % (j+1, i)) for j in range(n_vars)]
- # forecast sequence (t, t+1, ... t+n)
- for i in range(0, n_out):
- cols.append(df.shift(-i))
- if i == 0:
- names += [('var%d(t)' % (j+1)) for j in range(n_vars)]
- else:
- names += [('var%d(t+%d)' % (j+1, i)) for j in range(n_vars)]
- # put it all together
- agg = concat(cols, axis=1)
- agg.columns = names
- # drop rows with NaN values
- if dropnan:
- agg.dropna(inplace=True)
- return agg
-
- # load dataset
- dataset = read_csv('pollution.csv', header=0, index_col=0)
- values = dataset.values
- # integer encode direction
- encoder = LabelEncoder()
- values[:,4] = encoder.fit_transform(values[:,4])
- # ensure all data is float
- values = values.astype('float32')
- # normalize features
- scaler = MinMaxScaler(feature_range=(0, 1))
- scaled = scaler.fit_transform(values)
- # frame as supervised learning
- reframed = series_to_supervised(scaled, 1, 1)
- # drop columns we don't want to predict
- reframed.drop(reframed.columns[[9,10,11,12,13,14,15]], axis=1, inplace=True)
- print(reframed.head())
-
- # split into train and test sets
- values = reframed.values
- n_train_hours = 365 * 24
- train = values[:n_train_hours, :]
- test = values[n_train_hours:, :]
- # split into input and outputs
- train_X, train_y = train[:, :-1], train[:, -1]
- test_X, test_y = test[:, :-1], test[:, -1]
- # reshape input to be 3D [samples, timesteps, features]
- train_X = train_X.reshape((train_X.shape[0], 1, train_X.shape[1]))
- test_X = test_X.reshape((test_X.shape[0], 1, test_X.shape[1]))
- print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)
-
- # design network
- model = Sequential()
- model.add(LSTM(50, input_shape=(train_X.shape[1], train_X.shape[2])))
- model.add(Dense(1))
- model.compile(loss='mae', optimizer='adam')
- # fit network
- history = model.fit(train_X, train_y, epochs=50, batch_size=72, validation_data=(test_X, test_y), verbose=2, shuffle=False)
- # plot history
- pyplot.plot(history.history['loss'], label='train')
- pyplot.plot(history.history['val_loss'], label='test')
- pyplot.legend()
- pyplot.show()
-
- # make a prediction
- yhat = model.predict(test_X)
- test_X = test_X.reshape((test_X.shape[0], test_X.shape[2]))
- # invert scaling for forecast
- inv_yhat = concatenate((yhat, test_X[:, 1:]), axis=1)
- inv_yhat = scaler.inverse_transform(inv_yhat)
- inv_yhat = inv_yhat[:,0]
- # invert scaling for actual
- test_y = test_y.reshape((len(test_y), 1))
- inv_y = concatenate((test_y, test_X[:, 1:]), axis=1)
- inv_y = scaler.inverse_transform(inv_y)
- inv_y = inv_y[:,0]
- # calculate RMSE
- rmse = sqrt(mean_squared_error(inv_y, inv_yhat))
- print('Test RMSE: %.3f' % rmse)
运行以上代码,首先将会绘制训练过程中的训练和测试损失图。
训练中的每个epoch都会记录和绘制训练集和测试集的损失,并在整个训练结束后绘制模型最终的RMSE。下图中可以看到,整个模型的RMSE达到26.496。
- ...
- Epoch 46/50
- 0s - loss: 0.0143 - val_loss: 0.0133
- Epoch 47/50
- 0s - loss: 0.0143 - val_loss: 0.0133
- Epoch 48/50
- 0s - loss: 0.0144 - val_loss: 0.0133
- Epoch 49/50
- 0s - loss: 0.0143 - val_loss: 0.0133
- Epoch 50/50
- 0s - loss: 0.0144 - val_loss: 0.0133
- Test RMSE: 26.496
这个模型没有调优。你能做得更好吗?请在下面的评论中告诉我您的问题框架、模型配置和RMSE。
对于如何根据前面的多个时间步骤调整上面的示例来培训模型,已经有许多人提出了建议。在撰写最初的文章时,我尝试过这个方法和无数其他配置,但我决定不包含它们,因为它们没有提升模型技能。尽管如此,我在下面提供了这个示例作为参考模板,您可以根据自己的问题进行调整。在之前的多个时间步骤中训练模型所需的更改非常少,如下所示:首先,在调用series_to_supervised()时,必须适当地构造问题。我们将使用3小时的数据作为输入。还要注意,我们不再显式地从ob(t)的所有其他字段删除列。
- # specify the number of lag hours
- n_hours = 3
- n_features = 8
- # frame as supervised learning
- reframed = series_to_supervised(scaled, n_hours, 1)
接下来,在指定输入和输出列时需要更加小心。我们的框架数据集中有3 * 8 + 8列。我们将以3 * 8或24列作为前3小时所有特性的obs的输入。我们仅将污染变量作为下一小时的输出,如下所示:
- # split into input and outputs
- n_obs = n_hours * n_features
- train_X, train_y = train[:, :n_obs], train[:, -n_features]
- test_X, test_y = test[:, :n_obs], test[:, -n_features]
- print(train_X.shape, len(train_X), train_y.shape)
接下来,我们可以正确地重塑输入数据,以反映时间步骤和特征。
- # reshape input to be 3D [samples, timesteps, features]
- train_X = train_X.reshape((train_X.shape[0], n_hours, n_features))
- test_X = test_X.reshape((test_X.shape[0], n_hours, n_features))
模型拟合是一样的。唯一的另一个小变化是如何评估模型。具体来说,就是我们如何重构具有8列的行,这些行适合于反转缩放操作,从而将y和yhat返回到原始的缩放中,这样我们就可以计算RMSE。更改的要点是,我们将y或yhat列与测试数据集的最后7个特性连接起来,以便反向缩放,如下所示:
- # invert scaling for forecast
- inv_yhat = concatenate((yhat, test_X[:, -7:]), axis=1)
- inv_yhat = scaler.inverse_transform(inv_yhat)
- inv_yhat = inv_yhat[:,0]
- # invert scaling for actual
- test_y = test_y.reshape((len(test_y), 1))
- inv_y = concatenate((test_y, test_X[:, -7:]), axis=1)
- inv_y = scaler.inverse_transform(inv_y)
- inv_y = inv_y[:,0]
我们可以将所有这些修改与上面的示例联系在一起。多元时序多时滞输入预测的完整例子如下:
- from math import sqrt
- from numpy import concatenate
- from matplotlib import pyplot
- from pandas import read_csv
- from pandas import DataFrame
- from pandas import concat
- from sklearn.preprocessing import MinMaxScaler
- from sklearn.preprocessing import LabelEncoder
- from sklearn.metrics import mean_squared_error
- from keras.models import Sequential
- from keras.layers import Dense
- from keras.layers import LSTM
-
- # convert series to supervised learning
- def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):
- n_vars = 1 if type(data) is list else data.shape[1]
- df = DataFrame(data)
- cols, names = list(), list()
- # input sequence (t-n, ... t-1)
- for i in range(n_in, 0, -1):
- cols.append(df.shift(i))
- names += [('var%d(t-%d)' % (j+1, i)) for j in range(n_vars)]
- # forecast sequence (t, t+1, ... t+n)
- for i in range(0, n_out):
- cols.append(df.shift(-i))
- if i == 0:
- names += [('var%d(t)' % (j+1)) for j in range(n_vars)]
- else:
- names += [('var%d(t+%d)' % (j+1, i)) for j in range(n_vars)]
- # put it all together
- agg = concat(cols, axis=1)
- agg.columns = names
- # drop rows with NaN values
- if dropnan:
- agg.dropna(inplace=True)
- return agg
-
- # load dataset
- dataset = read_csv('pollution.csv', header=0, index_col=0)
- values = dataset.values
- # integer encode direction
- encoder = LabelEncoder()
- values[:,4] = encoder.fit_transform(values[:,4])
- # ensure all data is float
- values = values.astype('float32')
- # normalize features
- scaler = MinMaxScaler(feature_range=(0, 1))
- scaled = scaler.fit_transform(values)
- # specify the number of lag hours
- n_hours = 3
- n_features = 8
- # frame as supervised learning
- reframed = series_to_supervised(scaled, n_hours, 1)
- print(reframed.shape)
-
- # split into train and test sets
- values = reframed.values
- n_train_hours = 365 * 24
- train = values[:n_train_hours, :]
- test = values[n_train_hours:, :]
- # split into input and outputs
- n_obs = n_hours * n_features
- train_X, train_y = train[:, :n_obs], train[:, -n_features]
- test_X, test_y = test[:, :n_obs], test[:, -n_features]
- print(train_X.shape, len(train_X), train_y.shape)
- # reshape input to be 3D [samples, timesteps, features]
- train_X = train_X.reshape((train_X.shape[0], n_hours, n_features))
- test_X = test_X.reshape((test_X.shape[0], n_hours, n_features))
- print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)
-
- # design network
- model = Sequential()
- model.add(LSTM(50, input_shape=(train_X.shape[1], train_X.shape[2])))
- model.add(Dense(1))
- model.compile(loss='mae', optimizer='adam')
- # fit network
- history = model.fit(train_X, train_y, epochs=50, batch_size=72, validation_data=(test_X, test_y), verbose=2, shuffle=False)
- # plot history
- pyplot.plot(history.history['loss'], label='train')
- pyplot.plot(history.history['val_loss'], label='test')
- pyplot.legend()
- pyplot.show()
-
- # make a prediction
- yhat = model.predict(test_X)
- test_X = test_X.reshape((test_X.shape[0], n_hours*n_features))
- # invert scaling for forecast
- inv_yhat = concatenate((yhat, test_X[:, -7:]), axis=1)
- inv_yhat = scaler.inverse_transform(inv_yhat)
- inv_yhat = inv_yhat[:,0]
- # invert scaling for actual
- test_y = test_y.reshape((len(test_y), 1))
- inv_y = concatenate((test_y, test_X[:, -7:]), axis=1)
- inv_y = scaler.inverse_transform(inv_y)
- inv_y = inv_y[:,0]
- # calculate RMSE
- rmse = sqrt(mean_squared_error(inv_y, inv_yhat))
- print('Test RMSE: %.3f' % rmse)
拟合完成后输出如下:
- ...
- Epoch 45/50
- 1s - loss: 0.0143 - val_loss: 0.0154
- Epoch 46/50
- 1s - loss: 0.0143 - val_loss: 0.0148
- Epoch 47/50
- 1s - loss: 0.0143 - val_loss: 0.0152
- Epoch 48/50
- 1s - loss: 0.0143 - val_loss: 0.0151
- Epoch 49/50
- 1s - loss: 0.0143 - val_loss: 0.0152
- Epoch 50/50
- 1s - loss: 0.0144 - val_loss: 0.0149
训练集和测试集上的损失绘制如下图所示:
最后,RMSE测试被打印出来,并没有显示出任何技巧上的优势,至少在这个问题上没有。
Test RMSE: 27.177
我想补充一点,LSTM似乎不适用于自回归类型问题,您最好使用一个大窗口来研究MLP。我希望这个例子可以帮助您完成自己的时间序列预测实验。
赞 赏 作 者
Python中文社区作为一个去中心化的全球技术社区,以成为全球20万Python中文开发者的精神部落为愿景,目前覆盖各大主流媒体和协作平台,与阿里、腾讯、百度、微软、亚马逊、开源中国、CSDN等业界知名公司和技术社区建立了广泛的联系,拥有来自十多个国家和地区数万名登记会员,会员来自以工信部、清华大学、北京大学、北京邮电大学、中国人民银行、中科院、中金、华为、BAT、谷歌、微软等为代表的政府机关、科研单位、金融机构以及海内外知名公司,全平台近20万开发者关注。
推荐阅读:
▼点击阅读原文立即秒杀云产品!
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。