This example uses a fully connected neural network to classify tabular data representing different flowers, running purely in the browser using TensorFlow.js. The goal is to predict what kind of flower it is based on those features of each data point. The data comes from the famous Iris flower data set. The hyperparameters and optimizer used in training the network can be changed and weights learned can be visualised.
- The number of layers and number of neurons per layer can be changed and the architecture visualised. effect on the principal components generated.
- Effect of hyperparamters like learning rate and batch size on training can be compared.
- The oprimizer can be choosen from Adam RMSProp and SGD.
- The trained model can be tested on a user defined datal point.
- The weights of network can be visualised along with their magnitude.
To run it locally, you must install Python 3 or above and run the following command at the repository's root to set up a python server.
python -m http.serverYou can then browse to localhost:8000 and select index.html to view the application.
- Shashwat Gupta
- Sahil Goyal
- Ishan Kumar