avg_mae += hist.history['val_main_output_mean_absolute_error'][-1]
avg_loss += hist.history['val_main_output_loss'][-1]
avg_val_acc += hist.history['val_main_output_acc'][-1]
avg_mae_ske += hist.history['val_skeleton_output_mean_absolute_error'][
-1]
avg_loss_ske += hist.history['val_skeleton_output_loss'][-1]
avg_val_acc_ske += hist.history['val_skeleton_output_acc'][-1]
avg_mae_iner += hist.history[
'val_inertial_output_mean_absolute_error'][-1]
avg_loss_iner += hist.history['val_inertial_output_loss'][-1]
avg_val_acc_iner += hist.history['val_inertial_output_acc'][-1]
print("average mae : " + str(avg_mae / 5))
print("average loss : " + str(avg_loss / 5))
print("average accuracy: " + str(avg_val_acc / 5))
print("ske average mae: " + str(avg_mae_ske / 5))
print("ske average loss: " + str(avg_loss_ske / 5))
print("ske average accuracy: " + str(avg_val_acc_ske / 5))
print("iner average mae: " + str(avg_mae_iner / 5))
print("iner average loss: " + str(avg_loss_iner / 5))
print("iner average accuracy: " + str(avg_val_acc_iner / 5))
if __name__ == "__main__":
ROOT_DIR = os.path.dirname(os.path.abspath(__file__))
sys.path.append(ROOT_DIR + "/dataset")
trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner, trainY_iner, testX_iner, testY_iner = prepare_data(
)
run(trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner, trainY_iner,
testX_iner, testY_iner)
def main():
trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner, trainY_iner, testX_iner, testY_iner = prepare_data(
)
print("\n\n:::MLPC model:::")
main_nn_mlpc.run(trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner,
trainY_iner, testX_iner, testY_iner)
print("\n\n:::CNN model:::")
main_nn_cnn.run(trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner,
trainY_iner, testX_iner, testY_iner)
print("\n\n:::LSTM model:::")
main_lstm.run(trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner,
trainY_iner, testX_iner, testY_iner)
# print("\n\n:::LSTM ensemble model:::")
# main_lstm_ensemble.run(trainX_ske, trainY_ske, testX_ske, testY_ske, trainX_iner, trainY_iner, testX_iner,
# testY_iner)
print("\n\n:::Hybrid CNN + LSTM model:::")
main_cnn_lstm.run(trainX_ske, trainY_ske, testX_ske, testY_ske,
trainX_iner, trainY_iner, testX_iner, testY_iner)