Two Research Papers Accepted to Prestigious IEEE/ACIS SERA 2025 Conference
We are proud to announce that two research papers led by our students and faculty have been accepted for presentation at the 23rd IEEE/ACIS International Conference on Software Engineering, Management and Applications (SERA 2025), a prominent venue for interdisciplinary research in software engineering and applied computing.
1. XcepKNN: Leveraging Hybrid Deep Learning for Enhanced MRI-Based Brain Tumor Classification
Authors: Ethan Gilles, Yuqi Song, Xin Zhang, Fei Zuo
This paper highlights the exceptional work of Ethan Gilles, an undergraduate student and first author, who collaborated closely with faculty to address a critical challenge in medical diagnostics: accurate and early brain tumor detection. The research introduces XcepKNN, a novel deep learning architecture that integrates the Xception convolutional neural network with a K-Nearest Neighbor classifier to classify brain tumor types using MRI scans.
Tested on a dataset of 7,023 MRI images, the model achieved superior results compared to traditional methods, excelling in accuracy, precision, recall, and F1 score. The open-source implementation of XcepKNN provides a promising tool to support clinicians and researchers in improving diagnostic workflows for brain tumor detection.
2. Analyzing and Predicting Employee Turnover in the Restaurant Industry
Authors: Xin Zhang, Sarah Kayembe, Forest Ma, Yuqi Song, Fei Zuo
The second paper, which features Sarah Kayembe, a graduate student in our program, as the second author, results from a collaborative project with the Tourism and Hospitality program at USM. The study tackles the persistent issue of high employee turnover in the restaurant industry by applying machine learning (ML) and deep learning (DL) models to real-world survey data.
Using a dataset of 778 employee responses, the team explored various predictors of turnover, such as job satisfaction, compensation, and perceptions of management. Through models like Decision Trees, Random Forests, and Deep Neural Networks, the paper identifies actionable insights and predictive indicators to aid restaurants in developing effective employee retention strategies.
These recognitions at SERA 2025 underscore our department’s commitment to student-led innovation, interdisciplinary research, and real-world impact. Congratulations to all the authors on this significant achievement!