Skip to content
Emin Akif Erzurumlu
All work

Sismik Analiz

Can earthquake magnitude be predicted from its own past time series? My answer was no — and I am not hiding it.

Role
Deep learning model
Scope
Team project · 4 people
Year
2025
The problem

Earthquake records form a time series, and time-series forecasting has well-known architectures. The question was whether those architectures could predict the next magnitude from past magnitudes alone.

Approach
  • I split a 31,915-record dataset into 20-step windows.
  • I built two models in PyTorch: a bidirectional LSTM and an attention-based LSTM.
  • I measured success with R² rather than accuracy — accuracy is a meaningless metric when predicting a continuous value.
My part in this

My own notebook in this four-person team is PyTorch-based. The TensorFlow and Bi-GRU work in the team repository belongs to other members; I do not present it as mine.

Outcome

Bi-LSTM R² = 0.10, attention LSTM R² = 0.02. Both models fail to explain nearly any of the variance. Conclusion: a univariate setup cannot predict earthquake magnitude.

What I learned

Inflating these numbers would have been easy — overlap the windows and let train and test leak into each other, or swap the metric for accuracy. Either would have produced a handsome chart, and either would have been a lie. Showing that a model does not work is itself a finding; the real engineering is refusing to dress up a null result as a success.

Built with

Python · PyTorch · Jupyter · Bi-LSTM · Attention