Using machine learning to predict the time remaining until the next earthquake in lab. This is a google colab notebook.
This code was written to participate in a challenge hosted in caggle. The goal is to predict the time remaining until the next laboratory earthquake using real-time seismic data.
The data is provide by Los Alamos National Laboratory collected from the experimental setup used to study earthquake physics. The training data is a single, continuous segment of experimental data.
- Single training instance with various statistical features is generated by using 150000 Sequences. (except for LSTM).
- The different scikit-learn models and RNN is trained and the best one is used to predcit on test data.
- Evaluation metric: Mean Absolute Error
The notebook file is well commented and easy to understand, So I think I don't need to explain the whole process here.
Learn more about the data and competition here.