{"title":"重新想象线性探测:迁移学习中的柯尔莫哥洛夫-阿诺德网络","authors":"Sheng Shen, Rabih Younes","doi":"arxiv-2409.07763","DOIUrl":null,"url":null,"abstract":"This paper introduces Kolmogorov-Arnold Networks (KAN) as an enhancement to\nthe traditional linear probing method in transfer learning. Linear probing,\noften applied to the final layer of pre-trained models, is limited by its\ninability to model complex relationships in data. To address this, we propose\nsubstituting the linear probing layer with KAN, which leverages spline-based\nrepresentations to approximate intricate functions. In this study, we integrate\nKAN with a ResNet-50 model pre-trained on ImageNet and evaluate its performance\non the CIFAR-10 dataset. We perform a systematic hyperparameter search,\nfocusing on grid size and spline degree (k), to optimize KAN's flexibility and\naccuracy. Our results demonstrate that KAN consistently outperforms traditional\nlinear probing, achieving significant improvements in accuracy and\ngeneralization across a range of configurations. These findings indicate that\nKAN offers a more powerful and adaptable alternative to conventional linear\nprobing techniques in transfer learning.","PeriodicalId":501301,"journal":{"name":"arXiv - CS - Machine Learning","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Reimagining Linear Probing: Kolmogorov-Arnold Networks in Transfer Learning\",\"authors\":\"Sheng Shen, Rabih Younes\",\"doi\":\"arxiv-2409.07763\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper introduces Kolmogorov-Arnold Networks (KAN) as an enhancement to\\nthe traditional linear probing method in transfer learning. Linear probing,\\noften applied to the final layer of pre-trained models, is limited by its\\ninability to model complex relationships in data. To address this, we propose\\nsubstituting the linear probing layer with KAN, which leverages spline-based\\nrepresentations to approximate intricate functions. In this study, we integrate\\nKAN with a ResNet-50 model pre-trained on ImageNet and evaluate its performance\\non the CIFAR-10 dataset. We perform a systematic hyperparameter search,\\nfocusing on grid size and spline degree (k), to optimize KAN's flexibility and\\naccuracy. Our results demonstrate that KAN consistently outperforms traditional\\nlinear probing, achieving significant improvements in accuracy and\\ngeneralization across a range of configurations. These findings indicate that\\nKAN offers a more powerful and adaptable alternative to conventional linear\\nprobing techniques in transfer learning.\",\"PeriodicalId\":501301,\"journal\":{\"name\":\"arXiv - CS - Machine Learning\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Machine Learning\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.07763\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Machine Learning","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.07763","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
本文介绍了 Kolmogorov-Arnold 网络(KAN),作为迁移学习中传统线性探测方法的一种增强。线性探测通常应用于预训练模型的最后一层,但因其无法对数据中的复杂关系建模而受到限制。为了解决这个问题,我们建议用 KAN 代替线性探测层,KAN 利用基于样条的表示来逼近复杂的函数。在本研究中,我们将 KAN 与在 ImageNet 上预先训练好的 ResNet-50 模型进行了整合,并在 CIFAR-10 数据集上对其性能进行了评估。我们进行了系统的超参数搜索,重点关注网格大小和样条线度(k),以优化 KAN 的灵活性和准确性。我们的结果表明,KAN 的性能始终优于传统的线性探测,在各种配置下都能显著提高精度和泛化能力。这些发现表明,在迁移学习中,KAN 提供了一种比传统线性探测技术更强大、适应性更强的替代方法。
Reimagining Linear Probing: Kolmogorov-Arnold Networks in Transfer Learning
This paper introduces Kolmogorov-Arnold Networks (KAN) as an enhancement to
the traditional linear probing method in transfer learning. Linear probing,
often applied to the final layer of pre-trained models, is limited by its
inability to model complex relationships in data. To address this, we propose
substituting the linear probing layer with KAN, which leverages spline-based
representations to approximate intricate functions. In this study, we integrate
KAN with a ResNet-50 model pre-trained on ImageNet and evaluate its performance
on the CIFAR-10 dataset. We perform a systematic hyperparameter search,
focusing on grid size and spline degree (k), to optimize KAN's flexibility and
accuracy. Our results demonstrate that KAN consistently outperforms traditional
linear probing, achieving significant improvements in accuracy and
generalization across a range of configurations. These findings indicate that
KAN offers a more powerful and adaptable alternative to conventional linear
probing techniques in transfer learning.