用新型网络提升专家模型效率与可解释性，适用于金融与房产预测。
A Gated Residual Kolmogorov-Arnold Networks for Mixtures of Experts
- 用GRKAN替代传统门控机制，增强模型效率与可解释性。
- 在金融和房产数据上，性能优于传统MoE架构，尤其在序列任务中。
- 适合关注模型效率与可解释性的机器学习研究者使用。
本文提出KAMoE，一种基于门控残差柯尔莫哥洛夫-阿诺德网络（GRKAN）的新型混合专家（MoE）框架。我们以GRKAN作为传统门控函数的替代方案，旨在提升MoE建模中的效率与可解释性。在数字资产市场与房地产估值任务上的大量实验表明，KAMoE在多种任务与模型类型中均持续优于传统MoE架构。结果表明，相较于标准门控残差网络，GRKAN在基于LSTM的序列任务中表现更优。同时，我们分析了MoE与KAMoE架构中模型复杂度与性能提升之间的权衡关系。
原文摘要 · Abstract (English)
This paper introduces KAMoE, a novel Mixture of Experts (MoE) framework based on Gated Residual Kolmogorov-Arnold Networks (GRKAN). We propose GRKAN as an alternative to the traditional gating function, aiming to enhance efficiency and interpretability in MoE modeling. Through extensive experiments on digital asset markets and real estate valuation, we demonstrate that KAMoE consistently outperforms traditional MoE architectures across various tasks and model types. Our results show that GRKAN exhibits superior performance compared to standard Gating Residual Networks, particularly in LSTM-based models for sequential tasks. We also provide insights into the trade-offs between model complexity and performance gains in MoE and KAMoE architectures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。