In this tutorial, we will introduce how to tune neural network hyperparameters using grid search method in keras.

1. Import libraries

import numpy as np
from keras import models
from keras import layers
from keras.wrappers.scikit_learn import KerasClassifier
from sklearn.model_selection import GridSearchCV
from sklearn.datasets import make_classification

# Set random seed
np.random.seed(0)

2. Create data to train

# Number of features
number_of_features = 100

# Generate features matrix and target vector
features, target = make_classification(n_samples = 10000,
                                       n_features = number_of_features,
                                       n_informative = 3,
                                       n_redundant = 0,
                                       n_classes = 2,
                                       weights = [.5, .5],
                                       random_state = 0)

3. Build a neural network using keras

def create_network(optimizer='rmsprop'):
    
    # Start neural network
    network = models.Sequential()

    # Add fully connected layer with a ReLU activation function
    network.add(layers.Dense(units=16, activation='relu', input_shape=(number_of_features,)))

    # Add fully connected layer with a ReLU activation function
    network.add(layers.Dense(units=16, activation='relu'))

    # Add fully connected layer with a sigmoid activation function
    network.add(layers.Dense(units=1, activation='sigmoid'))

    # Compile neural network
    network.compile(loss='binary_crossentropy', # Cross-entropy
                    optimizer=optimizer, # Optimizer
                    metrics=['accuracy']) # Accuracy performance metric
    
    # Return compiled network
    return network

4. Wrap function In keras

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

5. Create a search space

# Create hyperparameter space
epochs = [5, 10]
batches = [5, 10, 100]
optimizers = ['rmsprop', 'adam']

# Create hyperparameter options
hyperparameters = dict(optimizer=optimizers, epochs=epochs, batch_size=batches)

6. Start grid search to find best hyperparameters

# Create grid search
grid = GridSearchCV(estimator=neural_network, cv=3, param_grid=hyperparameters)

# Fit grid search
grid_result = grid.fit(features, target)

7. Find best hyperparameters

# View hyperparameters of best neural network
grid_result.best_params_

Run this code, you may get the best hyperparameters:

{'batch_size': 10, 'epochs': 5, 'optimizer': 'adam'}