Pocket Neural Network aims to educate users about how neural networks work in an intuitive and interactive way. My goal is to also learn more about neural networks and machine learning in the process, sharing it with you all in the form of an app.
Current features:
* CNN / RNN (convolutional neural network, recurrent neural network)
* Binary Gate (2 inputs, 1 output, can imitate any gate)
* MNIST Exploration
* Adding layers to networks
* Creating new networks
* Visualizing layers and weights
* Tutorials and information guides on AI, NN, and ML.
To-Be-Determined / Backlog:
* Saving / Restoring neural networks and training models
* Sharing neural networks online or with friends
* Extend-able interface for activation layers and training data.
Current features:
* CNN / RNN (convolutional neural network, recurrent neural network)
* Binary Gate (2 inputs, 1 output, can imitate any gate)
* MNIST Exploration
* Adding layers to networks
* Creating new networks
* Visualizing layers and weights
* Tutorials and information guides on AI, NN, and ML.
To-Be-Determined / Backlog:
* Saving / Restoring neural networks and training models
* Sharing neural networks online or with friends
* Extend-able interface for activation layers and training data.
Show More
