Models Got Talent: Identifying High Performing Wearable Human Activity Recognition Models Without Training

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Goldman, Richard, Komperla, Varun, Ploetz, Thomas, Haresamudram, Harish
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908664688803840
author Goldman, Richard
Komperla, Varun
Ploetz, Thomas
Haresamudram, Harish
author_facet Goldman, Richard
Komperla, Varun
Ploetz, Thomas
Haresamudram, Harish
contents A promising alternative to the computationally expensive Neural Architecture Search (NAS) involves the development of Zero Cost Proxies (ZCPs), which correlate well with trained performance, but can be computed through a single forward/backward pass on a randomly sampled batch of data. In this paper, we investigate the effectiveness of ZCPs for HAR on six benchmark datasets, and demonstrate that they discover network architectures that obtain within 5% of performance attained by full-scale training involving 1500 randomly sampled architectures. This results in substantial computational savings as high-performing architectures can be discovered with minimal training. Our experiments not only introduce ZCPs to sensor-based HAR, but also demonstrate that they are robust to data noise, further showcasing their suitability for practical scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2511_06157
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Models Got Talent: Identifying High Performing Wearable Human Activity Recognition Models Without Training
Goldman, Richard
Komperla, Varun
Ploetz, Thomas
Haresamudram, Harish
Machine Learning
Artificial Intelligence
A promising alternative to the computationally expensive Neural Architecture Search (NAS) involves the development of Zero Cost Proxies (ZCPs), which correlate well with trained performance, but can be computed through a single forward/backward pass on a randomly sampled batch of data. In this paper, we investigate the effectiveness of ZCPs for HAR on six benchmark datasets, and demonstrate that they discover network architectures that obtain within 5% of performance attained by full-scale training involving 1500 randomly sampled architectures. This results in substantial computational savings as high-performing architectures can be discovered with minimal training. Our experiments not only introduce ZCPs to sensor-based HAR, but also demonstrate that they are robust to data noise, further showcasing their suitability for practical scenarios.
title Models Got Talent: Identifying High Performing Wearable Human Activity Recognition Models Without Training
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2511.06157