GluTANN: Transformer-Based Continuous Glucose Monitoring Model with ANN Attention
Sijie Xiong, Youhao Xu, Cheng Tang, Jianing Wang, Shuqing Liu, Atsushi Shimada

Project

• Abstract: In this work, we propose an innovative model based on Transformer architecture, GluTANN, with specially designed ANNs acting as self-attentions and paired correlations preserved by the encoder-decoder structure. Extensive experiments across five recognized datasets demonstrate that GluTANN has great competitiveness in reducing uncertainty while preserving satisfying accuracy, providing a feasible approach to effective glucose management and diabetes medical decisions.
• Core Idea: Reduce redundant tokens of Transformer-based models as much as we can.
• Domain: Time Series Forecasting, Diabetes, Glucose Monitoring

GluPIDHW: A PID-Holt-Winters Model for Personalized Glucose Monitoring
Youhao Xu+, Sijie Xiong+, Tao Sun, Jianing Wang, Cheng Tang, Atsushi Shimada

Project

• Abstract: The increasing worldwide incidence of diabetes has created a pressing demand for accurate and reliable glucose monitoring. Nevertheless, conventional machine learning methods exhibit limitations in trustworthiness and personalization, while deep learning methods with excellent trustworthiness underperform in accuracy and responsiveness. To realize an equilibrium, we propose an efficient model named GluPIDHW. Built upon an optimized Holt–Winters architecture augmented by a PID controller, GluPIDHW achieves improved accuracy and enhanced responsiveness. Extensive experiments conducted on five recognized datasets and out-of-distribution tasks demonstrate the superiority performance of GluPIDHW over excellent counterparts. This leadership is further validated by a Friedman testing and collectively, GluPIDHW offers a promising paradigm for continuous glucose monitoring and data-driven medical diagnose.
• Core Idea: Analogy from PID controller to Holt-Winters model.
• Domain: Time Series Forecasting, Diabetes, Glucose Monitoring

GlucoMixer: An Efficient Glucose Monitoring Model with Mixers
Sijie Xiong, Jianing Wang, Tao Sun, Cheng Tang, Fumiya Okubo, Atsushi Shimada

Project

• Abstract: With the global prevalence in diabetes and scarcity of definitive clinic schemes, the need for effective and reliable glucose monitoring has become imperative. Due to excellent responsiveness and precision, machine learning (ML) models have been widely employed by continuous glucose monitors (CGMs). However, recent studies have raised concerns that ML-based models often exhibit limited trustworthiness, injecting uncertainty into medical practices. In contrast, deep learning (DL) models are recognized as more trustworthy, but they are struggling with accuracy. To strike a balance between accuracy and trustworthiness, we propose GlucoMixer, an Encoder-only architecture built predominantly with Mixer modules. To prevent future information leakage, we design a Mask Block that employs a lightweight triangular masking scheme. Given the presence of two diabetes types, we employ a convolutional layer to distinguish relevant information and take two Time Mixer Blocks to handle distinct patterns accordingly. Comprehensive In-Distribution (ID) and Out-of-Distribution (OD) evaluations across five benchmark datasets highlight GlucoMixer’s superior performance in high predictive accuracy alongside excellent trustworthiness. According to the test ranking, GlucoMixer is more balanced across multiple evaluation metrics, demonstrating its potential as a practical spur for reliable glucose management and medical decisions.
• Core Idea: Combine linear projections and mask techniques to achieve a lightweight model.
• Domain: Time Series Forecasting, Diabetes, Glucose Monitoring

Enhancing Nonlinear Dependencies of Mamba via Negative Feedback for Time Series Forecasting
Sijie Xiong, Cheng Tang, Yuanyuan Zhang, Haoling Xiong, Youhao Xu, Atsushi Shimada

Project

• Abstract: In this work, we are inspired by the curvature from financial domains and control systems, proposing CME-Mamba. The effectiveness, stability, robustness, etc., are discussed. Extensive experiments demonstrate that CME-Mamba grows excellent to uncover complex paradigms and predict future states in various domains, especially improving the performance for periodic and high-variate situations.
• Core Idea: Leverage negative feedback loop to enhance non-linearity for TSF models.
• Domain: Time Series Forecasting, Control

KPMG: A Graphical Koopman-Mamba Approach for Financial Markets
Sijie Xiong, Cheng Tang, Fumiya Okubo, Tsubasa Minematsu, Yinlong Hu, Atsushi Shimada

Project

• Abstract: Financial markets reflected by indices are substantial components of global economy. While existing models have achieved significant forecasting performance, they struggle to balance temporal and variate dependencies, which results in a trade-off between predictive accuracy and trustworthiness. Moreover, current models suffer from disturbances and impurities embedded in the financial data. To address these challenges, we propose KPMG, an efficient architecture that integrates the strengths of Mamba and Graph Neural Networks. With crucial features emphasized by Koopman operator and both temporal and variate dependencies mixed up in KPMG, accuracy and trustworthiness are significantly advanced. Extensive experiments on nine leading benchmarks across three index datasets demonstrate that KPMG has superiority over counterparts in prediction performance, while remaining acceptable computational complexity. Ablation studies further confirm the effectiveness of each designed module. The Friedman test consolidates the superiority of KPMG over counterparts.
• Core Idea: Leverage Mamba, graphical neural network (GNN) to predict financial markets.
• Domain: Time Series Forecasting, Finance, Index Forecasting

Fine-tuned T5 Models on FairytaleQA Chinese Dataset
Sijie Xiong, Haoling Xiong, Tao Sun, Haiqiao Liu, Fumiya Okubo, Cheng Tang, Atsushi Shimada

Project

• Abstract: Question Answering (QA) is very important for comprehension learning and FairytaleQA is widely employed in this domain. However, rare versions in a limited number of alphabet languages restricts its application and current translators have five fatal errors. In our study, we manually translate FairytaleQA into Chinese and test its effectiveness via five fine-tuned T5 models.
• Core Idea: Extend current QA datasets on education for pre-trained models.
• Domain: Question-Answering, Education, Pre-trained Models