This is a Supervised learning neural networks project that detects the clothing material shown in the image format
MOTIVE : The key motive of this project was to learn about the working of neural networks
LIBRARIES USED:
Tensorflow
Keras
Matplotlib.pyplot(in case there is a need to plot the predicted images)
Numpy
USE :
This project can be used for educational purposes.
REQUIREMENTS :
A conda/virtualenv or any equivalent environment having installed the above libraries.
Working :
1.In this project we are using tensorflow as a backend and importing keras fashion mnist dataset that consists of grayscale images.
2.We then divide the dataset into two sets(a training and testing set).
3.A model is trained using the training set containing training images and training labels.
4.After training model the model is tested with the test data and the accuracy is noted.
5.The model is then saved for future uses.
Observations and recommendations:
1.It is observed that by increasing epochs rate the accuracy of the model can be increased but it comes at a cost of increase in training time.
2.It is also recommended to add a validation set in the system to improve its performance.
