AI-Based C++ Code Generation for Fitting Models in FITTER_WEB

8 Jul 2025, 17:00
15m
Room 310

Room 310

Speaker

Татьяна Соловьева (Jinr)

Description

The FITTER_WEB application is designed to solve the problem of fitting experimental data obtained in various physical experiments, and is deployed in the JINR cloud infrastructure. It uses the computing power of the ROOT package and provides a web interface for fitting data with theoretical models based on the resolution function. However, creating custom models required writing code manually, which could be difficult for some researchers.

This paper presents an extension of the functionality of this web application through integration with the DeepSeek artificial intelligence model. This allows you to automate the process of creating filtering functions in C++, which is especially useful for users who do not have deep programming knowledge.

Authors

Алексей Соловьев (JINR) Татьяна Соловьева (Jinr) Константин Лукьянов (JINR)

Presentation materials