Learning from Echoes

A Visual Journey into AI and the Dynamic Universe

This book explores artificial intelligence through the lens of astronomy, using images, light curves, spectra from light echoes to understand how AI works and what AI learns from our changing universe.

Preface¶

In this book

Foundations of AI¶

how machines learn from data

• Neural networks and feature learning

• Loss functions and optimization

• Evaluation and Interpretability

Learning from Space¶

how AI recognizes patterns in images

• Convolutional Neural Networks (CNNs)

• Feature extraction and classification

• Object detection

Generative Models¶

how AI creates new data

• autoencoders and VAE

• Segmentation and pixel-level understanding

• Generative Adversarial Networks (GANs)

• Diffusion model

Learning from Time¶

how AI models evolving systems

• Light curves and sequential data

• Recurrent Neural Networks (RNNs)

• Long Short-Term Memory networks (LSTMs)

• Transformers and attention mechanisms

• Predicting evolving astrophysical events

Multimodal Learning¶

how AI combines information across different forms of data

• Images, light curves, and spectra

• Cross-modal representations

• Joint learning across multiple data types

• Toward astronomical foundation models

• Future AI systems for next-generation sky surveys