Simple neural Network using Python and TensorFlow/Keras framework
Explanation
Step1: Install TensorFlow framework
Step2: Import necessary libraries
Step3: Load MNIST dataset and normalise pixel values between 1 and 0
Step4: Build Sequential Model
-Flatten: 28*28 input images in 1D array
-Dense(128,activation='relu'): connect 128 neurons and ReLU activation function
-dense(10): output layer with 10 neurons, using default linear activation in loss function
Step5: Complile model using Adam Optimizer, sparse categorical crossentropy loss function(for integer-encoded labels), and accuracy metrics
Step6: Train model for 10epochs, and validate on test data
Step7: Evaluate trained model on test data and evaluate accuracy