Skip to content

Latest commit

 

History

History
20 lines (14 loc) · 1.33 KB

README.md

File metadata and controls

20 lines (14 loc) · 1.33 KB

You can use the below citation: Nawal, Mayur, Kumar, Santosh, and Bharath Shekar. "LithoBot: An AutoML approach to identify lithofacies." Paper presented at the SEG/AAPG International Meeting for Applied Geoscience & Energy, Houston, Texas, USA, August 2022. doi: https://doi.org/10.1190/image2022-3751381.1

Lithobot

These are the codes and submitted paper for lithobot application. It is an Auto ML code to predict lithofacies from well logs.

Their are two .py files :

  1. training.py
  2. user_file.py

Details of the these files are as follows:

  1. In the file: training.py you will find the automatic data loading from force-2020 website, exploratory data analysis, data preprocessing, training of several models and predictions on test data.
  2. In the file: user_file.py user can upload their own data file and we will predict the output lithofacies based on their dataset. It will also give feedback on the dataset, so that user can consider getting more relevant features added to their data set, it also shows how much the results can improve if they follow the feedback.

Their is also .ipynb file

  1. lithobot_train.ipynb

Details of the these files are as follows:

  1. In this file, we have converted the complete training code into a google colab file so that users who wish to see the results in colaboratory file, can use it.