Search

Your search keyword '"Shen, Zhenqian"' showing total 10 results
Start Over Author "Shen, Zhenqian" Publication Year Range Last 3 years
10 results on '"Shen, Zhenqian"'

1. Benchmarking Graph Learning for Drug-Drug Interaction Prediction

Author

Shen, Zhenqian, Zhou, Mingyang, Zhang, Yongqi, and Yao, Quanming

Subjects
Computer Science - Machine Learning
Abstract

Predicting drug-drug interaction (DDI) plays an important role in pharmacology and healthcare for identifying potential adverse interactions and beneficial combination therapies between drug pairs. Recently, a flurry of graph learning methods have been introduced to predict drug-drug interactions. However, evaluating existing methods has several limitations, such as the absence of a unified comparison framework for DDI prediction methods, lack of assessments in meaningful real-world scenarios, and insufficient exploration of side information usage. In order to address these unresolved limitations in the literature, we propose a DDI prediction benchmark on graph learning. We first conduct unified evaluation comparison among existing methods. To meet realistic scenarios, we further evaluate the performance of different methods in settings with new drugs involved and examine the performance across different DDI types. Component analysis is conducted on the biomedical network to better utilize side information. Through this work, we hope to provide more insights for the problem of DDI prediction. Our implementation and data is open-sourced at https://anonymous.4open.science/r/DDI-Benchmark-ACD9/.

Published
2024

4. Classification of Visually Induced Motion Sickness Based on Phase-Locked Value Functional Connectivity Matrix and CNN-LSTM.

Author

Shen, Zhenqian, Liu, Xingru, Li, Wenqiang, Li, Xueyan, and Wang, Qiang

Subjects
*MOTION sickness, *FUNCTIONAL connectivity, *MACHINE learning, *CLASSIFICATION
Abstract

To effectively detect motion sickness induced by virtual reality environments, we developed a classification model specifically designed for visually induced motion sickness, employing a phase-locked value (PLV) functional connectivity matrix and a CNN-LSTM architecture. This model addresses the shortcomings of traditional machine learning algorithms, particularly their limited capability in handling nonlinear data. We constructed PLV-based functional connectivity matrices and network topology maps across six different frequency bands using EEG data from 25 participants. Our analysis indicated that visually induced motion sickness significantly alters the synchronization patterns in the EEG, especially affecting the frontal and temporal lobes. The functional connectivity matrix served as the input for our CNN-LSTM model, which was used to classify states of visually induced motion sickness. The model demonstrated superior performance over other methods, achieving the highest classification accuracy in the gamma frequency band. Specifically, it reached a maximum average accuracy of 99.56% in binary classification and 86.94% in ternary classification. These results underscore the model's enhanced classification effectiveness and stability, making it a valuable tool for aiding in the diagnosis of motion sickness. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

8. Efficient Hyper-parameter Optimization with Cubic Regularization

Author

Shen, Zhenqian, Yang, Hansi, Li, Yong, Kwok, Tin Yau, Yao, Quanming, Shen, Zhenqian, Yang, Hansi, Li, Yong, Kwok, Tin Yau, and Yao, Quanming

Abstract

As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning. In this paper, we consider a sub-class of hyper-parameter optimization problems, where the hyper-gradients are not available. Such problems frequently appear when the performance metric is non-differentiable or the hyper-parameter is not continuous. However, existing algorithms, like Bayesian optimization and reinforcement learning, often get trapped in local optimals with poor performance. To address the above limitations, we propose to use cubic regularization to accelerate convergence and avoid saddle points. First, we adopt stochastic relaxation, which allows obtaining gradient and Hessian information without hyper-gradients. Then, we exploit the rich curvature information by cubic regularization. Theoretically, we prove that the proposed method can converge to approximate second-order stationary points, and the convergence is also guaranteed when the lower-level problem is inexactly solved. Experiments on synthetic and real-world data demonstrate the effectiveness of our proposed method.

Published
2023

Catalog

Books, media, physical & digital resources
See catalog results