如何使用KerasClassifier将LSTM用于序列分类 [英] How to use LSTM for sequence classification using KerasClassifier

问题描述

我遇到一个binary classification问题，我需要根据2010-2015期间的客户互动来预测潜在的未来流行/流行产品.

I have a binary classification problem where I need to predict the potential future trendy/popular products based on customer interactions during 2010-2015.

当前，我的数据集包含1000 products，每个产品都标记为0或1(即二进制分类).标签是根据2016-2018期间的客户互动决定的.

Currently, my dataset includes 1000 products and each product is labelled as 0 or 1 (i.e. binary classification). The label was decided based on customer interactions during 2016-2018.

我正在计算2010-2015期间的how centrality measures changed over time for each product作为我的二进制分类问题的特征.例如，考虑下图，该图显示了每种产品degree centrality随时间的变化.

I am calculating how centrality measures changed over time for each product during 2010-2015 as the features for my binary classification problem. For example, consider the below figure that shows how degree centrality changed over time for each product.

更具体地说，我分析以下centrality measures的变化作为我的二进制分类问题的特征.

More specifically, I analyse the change of following centrality measures as the features for my binary classification problem.

• 每种商品的degree centrality在2010-2016年之间如何变化(请参见上图)
• 每种商品的betweenness centrality从2010-2016年如何变化
• 每种商品的closeness centrality从2010-2016年如何变化
• 每种商品的eigenvector centrality从2010-2016年如何变化

• how degree centrality of each good changed from 2010-2016 (see the above figure)
• how betweenness centrality of each good changed from 2010-2016
• how closeness centrality of each good changed from 2010-2016
• how eigenvector centrality of each good changed from 2010-2016

简而言之，我的数据如下所示.

In a nutshell, my data looks as follows.

product, change_of_degree_centrality, change_of_betweenness_centrality, change_of_closenss_centrality, change_of_eigenvector_centrality, Label
item_1, [1.2, 2.5, 3.7, 4.2, 5.6, 8.8], [8.8, 4.6, 3.2, 9.2, 7.8, 8.6], …, 1
item_2, [5.2, 4.5, 3.7, 2.2, 1.6, 0.8], [1.5, 0, 1.2, 1.9, 2.5, 1.2], …, 0
and so on.

我想使用深度学习模型来解决我的问题.在阅读教程时，我意识到LSTM适合我的问题.

I wanted to use deep learning model to solve my issue. When reading tutorials, I realised that LSTM suits my problem.

因此，我将使用下面提到的模型进行分类.

So, I am using the below mentioned model for my classification.

model = Sequential()
model.add(LSTM(10, input_shape=(6,4))) #where 6 is length of centrality sequence and 4 is types of centrality (i.e. degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality)
model.add(Dense(32))
model.add(Dense(1, activation=’sigmoid’))
model.compile(loss=’binary_crossentropy’, optimizer=’adam’, metrics=[‘accuracy’])

因为，我有一个小的数据集，我想执行10倍交叉验证.因此，我按照此教程.

Since, I have a small dataset I wanted to perform 10-fold cross-validation. So, I am using KerasClassifier as follows by following this tutorial.

print(features.shape) #(1000,6,4)
print(target.shape) #(1000) 

# Create function returning a compiled network
def create_network():
    model = Sequential()
    model.add(LSTM(10, input_shape=(6,4)))
    model.add(Dense(32))
    model.add(Dense(1, activation='sigmoid'))
    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])     

    return model

# Wrap Keras model so it can be used by scikit-learn
neural_network = KerasClassifier(build_fn=create_network, 
                                 epochs=10, 
                                 batch_size=100, 
                                 verbose=0)

print(cross_val_score(neural_network, features, target, cv=5))

但是，我指出在LSTM中使用cross validation是错误的(例如，本教程，此问题) .

However, I noted that it is wrong to use cross validation with LSTM (e.g., this tutorial, this question).

但是，我不清楚这是否适用于我，因为我只是在做 binary classification识别未来可能会流行/流行的产品(而不是预测产品).

However, I am not clear if this is applicable to me as I am only doing a binary classification predition to identify products that would be trendy/popular in future (not a forecasting).

我认为我的问题集中的数据在交叉验证中除以 point-wise ，而不是 time-wise .

I think the data in my problem setting is divided by point-wise in the cross-validation, but not time-wise.

即(逐点)

1st fold training:
item_1, item2, ........, item_799, item_800

1st fold testing:
item 801, ........, item_1000

不(按时间)

1st fold training:
2010, 2011, ........, 2015

1st fold testing:
2016, ........, 2018

由于这个事实，我认为在我的问题中使用cross validation是正确的.

Due to this fact, I am assuming that using cross validation is correct in my problem.

请让我知道根据我的问题和数据集使用交叉验证的合适方法.

Please let me know a suitable way to use cross-validation according to my problem and dataset.

注意:我不仅限于LSTM，也很高兴探索其他模型.

NOTE: I am not limited to LSTM and happy to explore other models as well.

如果需要，我很乐意提供更多详细信息.

I am happy to provide more details if needed.

推荐答案

交叉验证的类型很多，类似于神经网络的类型很多.在您的情况下，您尝试使用kfold交叉验证.

There are many types of cross validation similar to how there are many types of neural networks. In your case you are trying to use kfold cross validation.

在您链接的问题中，它正确地指出kfold交叉验证不应与时间序列数据一起使用.如果您正在对数据进行训练，然后对训练数据之前发生的数据进行测试，则无法准确评估模型.

In the question you linked, it correctly states that kfold cross validation should not be used with time series data. You can’t accurately evaluate your model if you are training on data and then testing on data that occurred before the training data.

但是，其他形式的交叉验证(例如提到的滑动窗口或扩展窗口)仍然可以与您的时间序列数据一起使用. sklearn中有一个使用扩展窗口方法拆分数据的功能. https://scikit-learn.org/stable/modules/生成/sklearn.model_selection.TimeSeriesSplit.html

However, other forms of cross validation (such as the mentioned sliding window or expanding window) will still work with your time series data. There is a function in sklearn that splits the data using the expanding window method. https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.TimeSeriesSplit.html

话虽如此，我不确定您是否真的在使用时间序列数据.如果只是简单地将每年的中心得分作为单独的功能，那么数据的顺序就无关紧要了，因为每一项只是一个数据点(假设一项的得分不会影响另一项).在这种情况下，您可以使用kfold交叉验证和其他可处理iid数据的网络.您甚至可以使用非神经网络，例如SVM或决策树.

With all that said, I am not sure if you are really using time series data. If you simply have the centrality scores for each year as a separate feature, then the order of your data does not matter since each item is only one data point (assuming that the scores of one item don’t impact another). In that case you can use kfold cross validation and other networks that work with iid data. You could even use non neural networks such as SVMs or decision trees.

这篇关于如何使用KerasClassifier将LSTM用于序列分类的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！