Artificial Neural Networks with Machine Learning

Artificial neural networks are one of the main tools used in machine learning. As the “neural” part of their name suggests, they are brain-inspired systems which are intended to replicate the way that we humans learn.

Neural networks consist of input and output layers, as well as (in most cases) a hidden layer consisting of units that transform the input into something that the output layer can use.

They are excellent tools for finding patterns which are far too complex or numerous for a human programmer to extract and teach the machine to recognize.

This Machine Learning Project Classifies Clothes from the Fashion MNIST Data set using Artificial Neural Networks and Python.

Let’s start by importing the libraries we need for this task

import tensorflow as tf
from tensorflow import keras
import numpy as np
import matplotlib.pyplot as plt

To load the data set

fashion = keras.datasets.fashion_mnist
(trainImages, trainLabels), (testImages, testLabels) = fashion.load_data()
imgIndex = 0
img = trainImages[imgIndex]
print("Image Label :",trainLabels[imgIndex])
plt.imshow(img)
#Output
Image Label : 9
<matplotlib.image.AxesImage at 0x7f1111a06d68>

To print the shape of the training and testing data

print(trainImages.shape)
print(testImages.shape)
#Output
(60000, 28, 28)
(10000, 28, 28)

Now let’s create a Neural Network

model = keras.Sequential([
                          keras.layers.Flatten(input_shape=(28,28)),
                          keras.layers.Dense(128, activation=tf.nn.relu),
                          keras.layers.Dense(10, activation=tf.nn.softmax)
])

To Compile the Model

model.compile(optimizer = 'adam',
           loss = 'sparse_categorical_crossentropy',
           metrics=['accuracy'])

To Train the model

model.fit(trainImages, trainLabels, epochs=5, batch_size=32)
#Output
Epoch 1/5
1875/1875 [==============================] - 4s 2ms/step - loss: 3.6150 - accuracy: 0.6802
Epoch 2/5
1875/1875 [==============================] - 4s 2ms/step - loss: 0.7296 - accuracy: 0.7488
Epoch 3/5
1875/1875 [==============================] - 4s 2ms/step - loss: 0.6374 - accuracy: 0.7725
Epoch 4/5
1875/1875 [==============================] - 4s 2ms/step - loss: 0.5873 - accuracy: 0.7906
Epoch 5/5
1875/1875 [==============================] - 4s 2ms/step - loss: 0.5579 - accuracy: 0.7993
<tensorflow.python.keras.callbacks.History at 0x7f1108dc3588>

To Evaluate the Model

model.evaluate(testImages, testLabels)
#Output
313/313 [==============================] - 0s 1ms/step - loss: 0.5916 - accuracy: 0.7981
[0.5915989279747009, 0.7980999946594238]

To Make a Prediction

predictions = model.predict(testImages[0:5])

# Print the predicted labels
print(predictions)
#Output
[[1.74235439e-07 2.69071290e-08 6.66509115e-20 3.09463957e-07
  1.11526007e-20 1.34603798e-01 8.10060641e-08 7.74199590e-02
  3.87958280e-05 7.87936807e-01]
 [2.89689321e-02 1.06601091e-02 6.28736615e-01 2.77338717e-02
  1.61624148e-01 1.49910515e-02 8.56256112e-02 1.23378839e-02
  2.35275514e-02 5.79410419e-03]
 [6.75366528e-06 9.99993205e-01 4.27281517e-12 2.68350314e-10
  8.65088672e-16 1.05001736e-14 1.33745196e-12 0.00000000e+00
  1.84386378e-11 0.00000000e+00]
 [6.56618613e-06 9.99993443e-01 1.46741508e-11 1.80866895e-08
  7.95811239e-14 1.56570215e-16 5.96713607e-12 0.00000000e+00
  3.94146077e-10 0.00000000e+00]
 [2.19924763e-01 1.00887669e-02 1.99720263e-01 6.23517819e-02
  4.97664846e-02 3.40277069e-07 4.30076748e-01 7.25772731e-09
  2.80708820e-02 2.27675168e-09]]

To print the maximum labels

print(np.argmax(predictions, axis=1))
# Print the actual label values
print(testLabels[0:5])
#Output
[9 2 1 1 6]
[9 2 1 1 6]

To Print the first 5 images

for i in range(0,5):
  plt.imshow(testImages[i], cmap='gray')
  plt.show()

Follow us on Instagram for all your Queries

Aman Kharwal
Aman Kharwal

I'm a writer and data scientist on a mission to educate others about the incredible power of data📈.

Articles: 1435

Leave a Reply