Learning with Analytical Models

TL;DR

This paper introduces a hybrid analytical and machine learning approach for performance modeling that reduces prediction costs and improves accuracy with small datasets, adaptable to hardware and workload variations.

Contribution

It presents a novel hybrid model that combines analytical and machine learning methods, enhancing performance prediction accuracy and efficiency.

Findings

Hybrid model outperforms pure machine learning in accuracy.

Model effectively learns and corrects analytical predictions.

Suitable for dynamic hardware and workload environments.

Abstract

To understand and predict the performance of scientific applications, several analytical and machine learning approaches have been proposed, each having its advantages and disadvantages. In this paper, we propose and validate a hybrid approach for performance modeling and prediction, which combines analytical and machine learning models. The proposed hybrid model aims to minimize prediction cost while providing reasonable prediction accuracy. Our validation results show that the hybrid model is able to learn and correct the analytical models to better match the actual performance. Furthermore, the proposed hybrid model improves the prediction accuracy in comparison to pure machine learning techniques while using small training datasets, thus making it suitable for hardware and workload changes.