基于矩阵乘积算符表示的序列化推荐模型
Matrix product operator based sequential recommendation model
刘沛羽 1姚博文 2高泽峰 3赵鑫1
作者信息
- 1. 中国人民大学高瓴人工智能学院,北京 100872
- 2. 中国人民大学物理系,北京 100872
- 3. 中国人民大学高瓴人工智能学院,北京 100872;中国人民大学物理系,北京 100872
- 折叠
摘要
推荐系统中的序列化推荐任务面临着高度复杂和多样性大的挑战,基于序列化数据的商品表示学习中广泛采用预训练和微调的方法,现有方法通常忽略了在新领域中模型微调可能会遇到的欠拟合和过拟合问题.为了应对这一问题,构建一种基于矩阵乘积算符(matrix product operator,MPO)表示的神经网络结构,并实现2 种灵活的微调策略.首先,通过仅更新部分参数的轻量化微调策略,有效地缓解微调过程中的过拟合问题;其次,通过增加可微调参数的过参数化微调策略,有力地应对微调中的欠拟合问题.经过实验验证,该方法在现有开源数据集上均实现显著的性能提升,充分展示在实现通用的物品表示问题上的有效性.
Abstract
The task of sequential recommendation confronts challenges characterized by high complexity and substantial diversity.The paradigm of pre-training and fine-tuning is extensively employed for learning item representations based on sequential data in recommendation scenarios.However,prevalent approaches tend to disregard the potential underfitting and overfitting issues that may arise during model fine-tuning in new domains.To address this concern,a novel neural network architecture grounded in the frame-work of matrix product operator(MPO)is introduced,and two versatile fine-tuning strategies are presented.Firstly,a lightweight fine-tuning approach that involves updating only a subset of parameters is proposed to effectively mitigate the problem of overfitting during the fine-tuning process.Secondly,an over-parameterization fine-tuning strategy is introduced by augmenting the number of trainable parameters,robustly addressing the issue of underfitting during fine-tuning.Through extensive experimentation on well-es-tablished open-source datasets,the efficacy of the proposed approach is demonstrated by achieving performance achievements.This serves as a compelling testament to the effectiveness of the proposed approach in addressing the challenge of general item representa-tion in recommendation systems.
关键词
推荐模型/序列化数据/矩阵乘积算符/过拟合/欠拟合Key words
recommendation model/sequential data/matrix product operator/overfitting/underfitting引用本文复制引用
基金项目
国家自然科学基金资助项目(62206299)
国家自然科学基金资助项目(62222215)
出版年
2024
NETL
NSTL
万方数据