Inception Neural Network for Complete Intersection Calabi-Yau 3-folds

TL;DR

This paper presents an Inception-inspired neural network that accurately predicts the Hodge number h^{1,1} of CICY 3-folds, significantly improving prediction accuracy with limited training data, thus aiding research in geometry and string theory.

Contribution

Introduces a novel neural network architecture based on Google's Inception model for predicting Hodge numbers of CICY 3-folds, achieving high accuracy with limited training data.

Findings

97% accuracy with 30% training data

99% accuracy with 80% training data

Neural networks effectively study geometric properties in mathematics and physics

Abstract

We introduce a neural network inspired by Google's Inception model to compute the Hodge number $h^{1,1}$ of complete intersection Calabi-Yau (CICY) 3-folds. This architecture improves largely the accuracy of the predictions over existing results, giving already 97% of accuracy with just 30% of the data for training. Moreover, accuracy climbs to 99% when using 80% of the data for training. This proves that neural networks are a valuable resource to study geometric aspects in both pure mathematics and string theory.