# fit and evaluate a model
def fit_evaluate_model(X_train, y_train, X_test, y_test):
    # build the model
    verbose, epochs, batch_size = 0, 10, 32
    n_timesteps, n_features, n_outputs = X_train.shape[1], X_train.shape[2], y_train.shape[1]
    # input head 1
    inputs1 = Input(shape=(n_timesteps,n_features))
    conv1 = Conv1D(filters=256, kernel_size=2, activation='relu')(inputs1)
    drop1 = Dropout(0.5)(conv1)
    pool1 = MaxPooling1D(pool_size=2)(drop1)
    flat1 = Flatten()(pool1)
    # input head 2
    inputs2 = Input(shape=(n_timesteps,n_features))
    conv2 = Conv1D(filters=128, kernel_size=3, activation='relu')(inputs2)
    drop2 = Dropout(0.3)(conv2)
    pool2 = MaxPooling1D(pool_size=2)(drop2)
    flat2 = Flatten()(pool2)
    # input head 3
    inputs3 = Input(shape=(n_timesteps,n_features))
    conv3 = Conv1D(filters=64, kernel_size=5, activation='relu')(inputs3)
    drop3 = Dropout(0.1)(conv3)
    pool3 = MaxPooling1D(pool_size=2)(drop3)
    flat3 = Flatten()(pool3)
    # concatenate
    merged = concatenate([flat1, flat2, flat3])
    # interpretation
    dense1 = Dense(100, activation='relu')(merged)
    outputs = Dense(n_outputs, activation='softmax')(dense1)
    model = Model(inputs=[inputs1, inputs2, inputs3], outputs=outputs)
    # save a plot of the model
    plot_model(model, show_shapes=True, to_file='multichannel.png')
    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
    # fit network
    model.fit([X_train,X_train,X_train], y_train, epochs=epochs, batch_size=batch_size, verbose=verbose)
    # evaluate model
    _, accuracy = model.evaluate([X_test,X_test,X_test], y_test, batch_size=batch_size, verbose=verbose)
    return accuracy
 
# summarize performance (accuracy, mean, sd)
def performance_summary(scores):
    print(scores)
    m, s = mean(scores), std(scores)
    print('Accuracy: %.3f%% (+/-%.3f)' % (m, s))
 
# run the trials
def run_trials(repeats=5):
    # repeat experiment
    scores = list()
    for repeat in range(repeats):
        score = fit_evaluate_model(X_train, y_train, X_test, y_test)
        score = score * 100.0
        print('>#%d: %.3f' % (repeat+1, score))
        scores.append(score)
    # performance summary 
    performance_summary(scores)
 
# run 
run_trials()