log_neural_net

1. An Artificial Neural Network (ANN) for synthetically generating sonic logs from conventional petro-physical logs, using a optimized neural network pipeline written in TensorFlow (Keras) - Python. The pipe-line is hyper-prametrically tuned.
2. For interacting with this Neural Network, a web-based Front-end, implemented in ReactJS has been provided.

Note:

The design and implementation of this ANN can be found at: ANN for Sonic Logs Prediction

Credits:

1. Paritosh Kumar Rakesh
2. Mayank Rakesh

ReactJS Based Front-end

Working Components

This whole set-up has 2 main components:

1. Front-End: ReactJS based - providing user interface for interacting with the Neural Network.
2. Back-end: Django Server hosting the Neural Network. The user selects the test file from from the UI in POST request and sends it to the back-end (django) for prediction. The django application running in the back-end loads the pre-trained ANN, receives the file from the POST request, cleans the data and feeds it to the loaded model. The prediction results are then returned in form of JSON.