nn_training_testing.py

import keras
import numpy as np
from os.path import isfile as is_file_exists

from keras.models import Sequential
from keras.layers import Dense, Activation
from keras.optimizers import Adam
from sklearn.preprocessing import MinMaxScaler
from sklearn.model_selection import train_test_split


def get_data(filename):
    _list = []
    with open(filename, "r") as f:
        for line in f.readlines():
            _list.append([int(line) for line in line.split(",")])
    return np.array(_list)

def get_label(filename):
    _list = []
    with open(filename, "r") as f:
        for label in f.readlines():
            _list.append(int(label))
    return np.array(_list)


model = Sequential()

# Layers
model.add(Dense(24,activation='relu'))
model.add(Dense(7,activation='softmax'))

model.compile(optimizer='adam',
              loss='categorical_crossentropy',
              metrics=['accuracy'],
              batch_size=16)

data = get_data("data.txt")
labels = get_label("data_labels.txt")

X_train, X_test, y_train, y_test = train_test_split(data, labels,
                                                    test_size=0.40, random_state=42)
training_labels = keras.utils.to_categorical(y_train, num_classes=7)
testing_labels = keras.utils.to_categorical(y_test, num_classes=7)

if not is_file_exists("trained_model"):
    model.fit(X_train,training_labels,epochs=7)
    model.save('trained_model')
else:
    model = keras.models.load_model('trained_model')

test_loss, test_accuracy = model.evaluate(X_test, testing_labels)
print("Test Loss", test_loss)
print("Test Accuracy", test_accuracy)