README.MNIST

DEEP LEARNING PROJECT- MASTERS COURSE WORK

1.Here we are taking the MNIST fashion data set and using tensorflow libraries.

Tensorflow used for: It is more powerful than the sequential API in the sense branching or sharing of layers is allowed here. And also it can have multiple inputs and outputs

Dataset: Fashion-MNIST is a dataset of Zalando's article images—consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. Zalando intends Fashion-MNIST to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms. It shares the same image size and structure of training and testing splits.

2.There are two methods in training the DEEP LEARNING MODEL ,those are sequential method and functional api

3.In sequential , we cannot have multiple input at a time , it is designed layer by layer

4.First we did batchnormalization and dropout layering for sequential methods, we got accuracy of 86 and 90 % respectively,

5.Functional api : It is more advanced than the sequential API in the sense branching or sharing of layers is allowed here and also it can have multiple inputs and outputs.

Optimizer:Here we used Adam optimizer, it is a replacement optimization algorithm for stochastic gradient descent for training deep learning models.

6.After adding the batchnormaliation and drop out layers , we compile the model and fit the model to run epochs with set limit.

AActivation function: it defines the what node/nueron has to do .

7.So after the fit we got 90.8 % and 87% accuracy in functional api.