Machine Learning for Seismic Data

Where learned models genuinely help in seismic workflows and where they do not. Covers CNN denoising, interpolation, fault detection with synthetic training pairs, and physics-informed networks, each with runnable examples in the browser.

17 interactive sections across 8 books
Start here → ML in processing: where it fits, from Seismic Data Processing

Seismic Data Processing

  1. ML in processing: where it fits
  2. Denoising with CNNs
  3. Interpolation / reconstruction
  4. ML-assisted first-break picking

Synthetic Seismic Modeling

  1. Fault-Detection Training Pairs
  2. Fault-Detection Dataset

Applied Machine Learning for Geoscientists

  1. Deep Learning: Convolutional Neural Networks
  2. CNN Practice
  3. Building a Simple Neural Network
  4. Recurrent Neural Networks
  5. Gradient Descent and Sampling Strategies

3D Seismic Interpretation

  1. Machine-learning QI: neural networks for facies classification
  2. Reading inversion products: elastic to rock properties

Statistics and Data Science for Researchers

  1. Calibration: when 95% really means 95%

A Mathematician's Survey

  1. Why Linearity Is Universal

Geostatistics

  1. Generative models for geostatistics (GANs, diffusion)

Physics-Informed Neural Networks for Exploration Seismology

  1. Training loops and optimisers

This page is prerendered for SEO and accessibility. Every linked section hydrates live interactive widgets on JavaScript load.