# NGO-GM: Natural Gradient Optimization for Graphical Models

## TL;DR

This paper introduces NGO-GM, a natural gradient optimization method for graphical models that improves parameter learning by directly optimizing the objective function, outperforming traditional methods like EM.

## Contribution

It presents a novel natural gradient-based approach for parameter estimation in graphical models, offering a more direct and potentially more effective alternative to EM.

## Key findings

- Better performance than traditional methods in trend detection
- Less prone to overfitting in financial market analysis
- Theoretically justified and empirically validated approach

## Abstract

This paper deals with estimating model parameters in graphical models. We reformulate it as an information geometric optimization problem and introduce a natural gradient descent strategy that incorporates additional meta parameters. We show that our approach is a strong alternative to the celebrated EM approach for learning in graphical models. Actually, our natural gradient based strategy leads to learning optimal parameters for the final objective function without artificially trying to fit a distribution that may not correspond to the real one. We support our theoretical findings with the question of trend detection in financial markets and show that the learned model performs better than traditional practitioner methods and is less prone to overfitting.