Evaluate the Performance Of Deep Learning Models in Keras (using different techniques)

Data Splitting

The large amount of data and the complexity of the models require very long training times.

As such, it is typically to use a simple separation of data into training and test datasets or training and validation datasets.

Keras provides a two convenient ways of evaluating your deep learning algorithms this way:

1. Use an automatic verification dataset.
2. Use a manual verification dataset.

Use a Automatic Verification Dataset

Keras can separate a portion of your training data into a validation dataset and evaluate the performance of your model on that validation dataset each epoch.

You can do this by setting the validation_split argument on the fit() function to a percentage of the size of your training dataset.

For example, a reasonable value might be 0.2 or 0.33 for 20% or 33% of your training data held back for validation.

The example below demonstrates the use of using an automatic validation dataset on a small binary classification problem. All examples in this post use the Pima Indians onset of diabetes dataset. You can download it from the UCI Machine Learning Repository and save the data file in your current working directory with the filename pima-indians-diabetes.csv (update: download from here).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

# MLP with automatic validation set from keras.models import Sequential from keras.layers import Dense import numpy # fix random seed for reproducibility numpy.random.seed(7) # load pima indians dataset dataset = numpy.loadtxt("pima-indians-diabetes.csv", delimiter=",") # split into input (X) and output (Y) variables X = dataset[:,0:8] Y = dataset[:,8] # create model model = Sequential() model.add(Dense(12, input_dim=8, activation='relu')) model.add(Dense(8, activation='relu')) model.add(Dense(1, activation='sigmoid')) # Compile model model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']) # Fit the model model.fit(X, Y, validation_split=0.33, epochs=150, batch_size=10)

Running the example, you can see that the verbose output on each epoch shows the loss and accuracy on both the training dataset and the validation dataset.

Use a Manual Verification Dataset

Keras also allows you to manually specify the dataset to use for validation during training.

In this example we use the handy train_test_split() function from the Python scikit-learn machine learning library to separate our data into a training and test dataset. We use 67% for training and the remaining 33% of the data for validation.

The validation dataset can be specified to the fit() function in Keras by the validation_data argument. It takes a tuple of the input and output datasets.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

# MLP with manual validation set from keras.models import Sequential from keras.layers import Dense from sklearn.model_selection import train_test_split import numpy # fix random seed for reproducibility seed = 7 numpy.random.seed(seed) # load pima indians dataset dataset = numpy.loadtxt("pima-indians-diabetes.csv", delimiter=",") # split into input (X) and output (Y) variables X = dataset[:,0:8] Y = dataset[:,8] # split into 67% for train and 33% for test X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.33, random_state=seed) # create model model = Sequential() model.add(Dense(12, input_dim=8, activation='relu')) model.add(Dense(8, activation='relu')) model.add(Dense(1, activation='sigmoid')) # Compile model model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']) # Fit the model model.fit(X_train, y_train, validation_data=(X_test,y_test), epochs=150, batch_size=10)

Like before, running the example provides verbose output of training that includes the loss and accuracy of the model on both the training and validation datasets for each epoch.

Manual k-Fold Cross Validation

The gold standard for machine learning model evaluation is k-fold cross validation.

It provides a robust estimate of the performance of a model on unseen data. It does this by splitting the training dataset into k subsets and takes turns training models on all subsets except one which is held out, and evaluating model performance on the held out validation dataset. The process is repeated until all subsets are given an opportunity to be the held out validation set. The performance measure is then averaged across all models that are created.

Cross validation is often not used for evaluating deep learning models because of the greater computational expense. For example k-fold cross validation is often used with 5 or 10 folds. As such, 5 or 10 models must be constructed and evaluated, greatly adding to the evaluation time of a model.

Nevertheless, it when the problem is small enough or if you have sufficient compute resources, k-fold cross validation can give you a less biased estimate of the performance of your model.

In the example below we use the handy StratifiedKFold class from the scikit-learn Python machine learning library to split up the training dataset into 10 folds. The folds are stratified, meaning that the algorithm attempts to balance the number of instances of each class in each fold.

The example creates and evaluates 10 models using the 10 splits of the data and collects all of the scores. The verbose output for each epoch is turned off by passing verbose=0 to the fit() and evaluate() functions on the model.

The performance is printed for each model and it is stored. The average and standard deviation of the model performance is then printed at the end of the run to provide a robust estimate of model accuracy.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

# MLP for Pima Indians Dataset with 10-fold cross validation from keras.models import Sequential from keras.layers import Dense from sklearn.model_selection import StratifiedKFold import numpy # fix random seed for reproducibility seed = 7 numpy.random.seed(seed) # load pima indians dataset dataset = numpy.loadtxt("pima-indians-diabetes.csv", delimiter=",") # split into input (X) and output (Y) variables X = dataset[:,0:8] Y = dataset[:,8] # define 10-fold cross validation test harness kfold = StratifiedKFold(n_splits=10, shuffle=True, random_state=seed) cvscores = [] for train, test in kfold.split(X, Y):   # create model model = Sequential() model.add(Dense(12, input_dim=8, activation='relu')) model.add(Dense(8, activation='relu')) model.add(Dense(1, activation='sigmoid')) # Compile model model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']) # Fit the model model.fit(X[train], Y[train], epochs=150, batch_size=10, verbose=0) # evaluate the model scores = model.evaluate(X[test], Y[test], verbose=0) print("%s: %.2f%%" % (model.metrics_names[1], scores[1]*100)) cvscores.append(scores[1] * 100) print("%.2f%% (+/- %.2f%%)" % (numpy.mean(cvscores), numpy.std(cvscores)))

Running the example will take less than a minute and will produce the following output:

1 2 3 4 5 6 7 8 9 10 11

acc: 77.92% acc: 68.83% acc: 72.73% acc: 64.94% acc: 77.92% acc: 35.06% acc: 74.03% acc: 68.83% acc: 34.21% acc: 72.37% 64.68% (+/- 15.50%)

Summary

In this post you discovered the importance of having a robust way to estimate the performance of your deep learning models on unseen data.

You discovered three ways that you can estimate the performance of your deep learning models in Python using the Keras library:

• Use Automatic Verification Datasets.
• Use Manual Verification Datasets.
• Use Manual k-Fold Cross Validation.