Multi-class classification for Fashion-MNIST in tensorflow
Assignment 3 code for Deep Learning, CS60010.
MNIST data provides us very high accuracy with simple models, so we will be using fashion-MNIST.
The neural network has 3 hidden layers, with 50 epochs/iterations. Refer to Report for more details.
-
Training accuracy: 96.83% (Max accuracy in an iteration: 100%)
-
Testing accuracy: 89.46%
-
Loss as a function of iterations
- Accuracy as a function of iterations
We apply Logistic Regression at every hidden layer. Here are the results:
- Layer 1: 88.87%
- Layer 2: 89.33%
- Layer 3: 89.46%
The first layer seems to provide enough accuracy, which proves further layers might not be needed.
python train.py --train
Run training, save weights into weights/ folder.
python train.py --train iter=5
Run training with specified number of iterations. Default iterations are 50.
python train.py --test
Load precomputed weights and report test accuracy.
python train.py --layer=1
Run Logistic Regression on hidden layer's output and report the accuracy. Allowed options : 1, 2, 3.
data_loaderis used to load data from zip files indatafolder.moduledefines the neural network parameters, and network related code.trainhandles input and states the model.
The MIT License (MIT) 2018 - Kaustubh Hiware.