Saved in:
Bibliographic Details
Main Authors: Chan, Alena, Garmonina, Maria
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.04707
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908480737116160
author Chan, Alena
Garmonina, Maria
author_facet Chan, Alena
Garmonina, Maria
contents We evaluate the performance of several optimizers on the task of forecasting S&P 500 Index returns with the MambaStock model. Among the most widely used algorithms, gradient-smoothing and adaptive-rate optimizers (for example, Adam and RMSProp) yield the lowest test errors. In contrast, the Lion optimizer offers notably faster training. To combine these advantages, we introduce a novel family of optimizers, Roaree, that dampens the oscillatory loss behavior often seen with Lion while preserving its training speed.
format Preprint
id arxiv_https___arxiv_org_abs_2508_04707
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle From Rattle to Roar: Optimizer Showdown for MambaStock on S&P 500
Chan, Alena
Garmonina, Maria
Computational Finance
Machine Learning
We evaluate the performance of several optimizers on the task of forecasting S&P 500 Index returns with the MambaStock model. Among the most widely used algorithms, gradient-smoothing and adaptive-rate optimizers (for example, Adam and RMSProp) yield the lowest test errors. In contrast, the Lion optimizer offers notably faster training. To combine these advantages, we introduce a novel family of optimizers, Roaree, that dampens the oscillatory loss behavior often seen with Lion while preserving its training speed.
title From Rattle to Roar: Optimizer Showdown for MambaStock on S&P 500
topic Computational Finance
Machine Learning
url https://arxiv.org/abs/2508.04707