Team ID: 12
Vineetha Gummadi, Class ID: 10

• To implement the MNIST dataset and clothes data set using Zalando Fashion MNIST dataset.
• To change the different parameters and calculating the training and test accuracy of MNIST and fashion dataset.

Downloaded and loaded the data

Trained the model and tested the model

Below are the observations for certain parameters

Fashion dataset: Adam Optimizer
Layer 1 filter size: 32
Layer 2 filter size: 64
Final layer: 1028
test accuracy 0.9463

Below are the observations

Fashion dataset: Adam optimizer
Layer 1 filter size: 16
Layer 2 filter size: 36
Final layer: 128
test accuracy 0.8994

Fashion dataset: Gradient Descent optimizer
Layer 1 filter size: 32
Layer 2 filter size: 64
Final layer: 1028
test accuracy 0.4087

Fashion dataset: Gradient Descent optimizer 
Layer 1 filter size: 16
Layer 2 filter size: 36
Final layer: 128
test accuracy 0.2079

MNIST dataset: Adam Optimizer
Layer 1 filter size: 32
Layer 2 filter size: 64
Final layer: 1028
test accuracy 0.9342

MNIST dataset: Adam optimizer
Layer 1 filter size: 16
Layer 2 filter size: 36
Final layer: 128
test accuracy 0.908

MNIST dataset: Gradient Descent optimizer
Layer 1 filter size: 32
Layer 2 filter size: 64
Final layer: 1028
test accuracy 0.3755

MNIST dataset: Gradient Descent optimizer 
Layer 1 filter size: 16
Layer 2 filter size: 36
Final layer: 128
test accuracy 0.1146