IEEE ICDM 2025: Undergraduate and High School Symposium

Home

We are delighted to announce the 2nd Undergraduate and High School Symposium, to be held as part of the IEEE ICDM 2025 Conference. This symposium aims to provide a platform for young researchers to showcase their innovative work in the field of data mining and related disciplines.

Why Participate in this Symposium?

Present your research to an international audience at one of the premier data mining conferences.
Compete for Outstanding Paper Awards, Runner-Up Awards, and Rising Star Awards honoring exceptional student contributions.
Interact directly with top AI researchers, industry innovators, and NSF Program Directors.

Program Agenda

Nov 14: conference banquet (evening). Nov 15: full-day UGHS program.

November 14, 2025 — Conference Banquet
Time	Event	Location
18:00–21:00	Banquet	—

November 15, 2025 (Location: 1 main room + 5 additional parallel session rooms)
Time	Event	Speaker
09:00–10:00	Keynote (50 mins) Location: Senate Room	NSF Program Director (TBD)
10:00–10:30	Coffee Break Location: Capital Terrace, Congressional	—
10:30–11:25	Fireside Chat (50 mins) with Dr. Jure Leskovec (Stanford) Location: Senate Room	Moderator: TBD
11:30–12:25	Panel with Parents (50 mins): Next Generation AI Education Location: Senate Room	Panelists: See panelists →
12:25–12:30	Poster Preparation	—
12:30–14:00	Poster Session (Lunch) Location: Capital Terrace, Congressional	All Authors
14:00–15:30	Oral Presentations (12 mins × 7 papers × 6 rooms) UH1 - 6 · Locations see details	Oral Sessions: See sessions →
15:30–16:00	Coffee Break Location: Capital Terrace, Congressional	—
16:00–17:30	Oral Presentations (12 mins × 7 papers × 6 rooms) UH7 - 12 · Locations see details	Oral Sessions: See sessions →
17:30	Closing (Award Announcements) Location: Presidential Room	—

Oral Presentations · 14:00–15:30

← Back to agenda

UH1 — Federal A UH2 — Federal B UH3 — South American B UH4 — California UH5 — New York UH6 — Massachusetts

14:00–15:30 · Oral Presentations（12 mins × 7 papers × 6 rooms）

Session Chairs — UH1–UH6

UH1

Sanjay Madria

UH2

Chen Chen

UH3

Fang Jin

UH4

Haibing Lu

UH5

Mohammad Ali Javidian

UH6

Long Nguyen

UH1 (Room: Federal A, Session Chair: Sanjay Madria): S01287, S01286, S01285, S01242, S01241, S01282, S01281

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01287`	Can Small Quantized VLMs Drive? An Experimental Evaluation of Small Quantized VLMs for Autonomous Driving	Samson Mathew
2	14:12–14:24	`S01286`	Data Insights into Teen Consumer Trends: From Kaggle to Knoxville	Chuanren Liu
3	14:24–14:36	`S01285`	An Empirical Approach Toward Understanding the Impact of Essential Oils on Alzheimer's Disease Progression	Khalil Al-Hussaeni
4	14:36–14:48	`S01242`	Assessing Bias Within Diabetes Risk Prediction in Machine Learning Techniques	Ayesha Faruki
5	14:48–15:00	`S01241`	Heatwaves and Health Risks in New York City	Jie Yang
6	15:00–15:12	`S01282`	Interpretable Feature Mining for AI Product Design	Vivian Foutz
7	15:12–15:24	`S01281`	QuizWhiz: An End-to-End AI-Powered Educational Platform for K-12 Intelligent Tutoring and Teaching Analytics	Ming Zhang

UH2 (Room: Federal B, Session Chair: Chen Chen): S01280, S01279, S01278, S01277, S01276, S01275, S01274

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01280`	Early Wildfire Detection with UAVs using a Frame Difference Method	Brian Hong
2	14:12–14:24	`S01279`	FS-PREM: A Physics-Aware Framework for Predicting Port Disruption	Shriraghav Ashok
3	14:24–14:36	`S01278`	Privacy-First Triage Classification with Open-Weight LLMs: A Chain-of-Thought Distillation Approach	Zeyuan Zhao
4	14:36–14:48	`S01277`	DETECT: Data-Driven Evaluation of Treatments Enabled by Classification Transformers	Yuanheng Mao
5	14:48–15:00	`S01276`	SmartPharynx: A Camera-Based Smartphone System for Screening of Bacterial Pharyngitis with a Low-Shot CycleGAN and Custom CNN	Srikar Kovvali
6	15:00–15:12	`S01275`	An Analysis of Gender-Based Differential Item Functioning in the PISA 2018 and 2022 Cycles	Isabel Xiong
7	15:12–15:24	`S01274`	A Critical Analysis of a Multi-Input CNN Architecture for Quantum-Enhanced Forecasting	Ateef Mahmud

UH3 (Room: South American B, Session Chair: Fang Jin): S01273, S01271, S01269, S01267, S01264, S01263

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01273`	Towards Robust Anomaly Detection in Fish Behavior: Hybrid LLM–ML Ensembles and Federated Learning	Qiong Cheng
2	14:12–14:24	`S01271`	Transforming Color Correction for Colorblindness with Hydrodynamic Modeling and Deep Learning-Based Validation	Lucas Yang
3	14:24–14:36	`S01269`	Comparative Analysis of GraphCast and the Global Forecast System Using Real-Time Mesoscale Analysis	Jonathan Yu
4	14:36–14:48	`S01267`	Learning for Inflation Forecasting with Dynamic Feature Spaces	Zakariyya Scavotto
5	14:48–15:00	`S01264`	Predicting Residual Cognitive Deficit Post-Ischemic Stroke: An Imbalance-Aware Machine Learning Pipeline on EHR Data	Sirichandana Yakkala
6	15:00–15:12	`S01263`	A ResNet and ViT-U-Net Hybrid Model for Accurate FVM Flooding Simulations	Isabella Cho

UH4 (Room: California, Session Chair: Haibing Lu): S01262, S01261, S01260, S01258, S01257, S01256

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01262`	AI-Assisted Safe Drop Zone Identification for Human-Guided Drone Deliveries	Atharva Kakatkar
2	14:12–14:24	`S01261`	Adaptive Execution Scheduler for DataDios SmartDiff	Aryan Poduri
3	14:24–14:36	`S01260`	Contrastive Retrieval Augmented In-Context Learning for Medical Classification Tasks with Imbalanced Data	Swarnika Joshi
4	14:36–14:48	`S01258`	Graph Perspective on Multi-modal Mouse Neural Data and Behavior Analysis	Wenhao Hu
5	14:48–15:00	`S01257`	Efficient Semantic-based Video Segment Querying	ziqi Zhou
6	15:00–15:12	`S01256`	Enhancing Radiographic Disease Detection with MetaCheX, a Context-Aware Multimodal Model	Nathan He

UH5 (Room: New York, Session Chair: Mohammad Ali Javidian): S01255, S01254, S01253, S01252, S01251, S01250

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01255`	Statistical Mining of Patient Reviews for Geographic Insights on Quality of Urological Care: A Pilot Study	Max Yu
2	14:12–14:24	`S01254`	Enhancing On-Chip Learning for RRAM Devices Through Evolutionary Theories	Xinghui Zhao
3	14:24–14:36	`S01253`	Application of Object Segmentation Model in Contact Angle Measurement for Hydrophobicity Determination	Joann Xie
4	14:36–14:48	`S01252`	Hybrid BiLSTM-RF Framework for Lithium-ion Battery State of Health and RUL Prediction	Irene Lu
5	14:48–15:00	`S01251`	Uncertainty Quantification in Deep Learning based Breast Cancer Diagnosis using DCE-MRI and TRAMs	Jerry Wang
6	15:00–15:12	`S01250`	MetaRef: A Generalizable Physics-Aware Refinement Framework for Metamaterial Design	Alexander Lu

UH6 (Room: Massachusetts, Session Chair: Long Nguyen): S01249, S01248, S01246, S01245, S01244, S01243, DM556

#	Time	Paper ID	Title	Contact
1	14:00–14:12	`S01249`	Few-Shot Learning Meets Large Language Models: Mining Medicine Interventions From Reddit	Fang Jin
2	14:12–14:24	`S01248`	Assessing Cognitive Biases in LLMs for Judicial Decision Support: Virtuous Victim and Halo Effects	Sierra Liu
3	14:24–14:36	`S01246`	FinFraud-LLM: Exploring Large Language Models for Financial Fraud Detection	Johnson Chen
4	14:36–14:48	`S01245`	Can Reasoning LLMs Eliminate Conformity in Multi-Agent Systems?	Alina Liu
5	14:48–15:00	`S01244`	Echo State Networks in Reservoir Computing: Foundations, Benchmarks, and Applications to Next-G Wireless Communication	Andrew Liu
6	15:00–15:12	`S01243`	Interpretable Deep Learning Framework for the Diagnosis of Age-Related Macular Degeneration	Anvitaa Rudharraju
7	15:12–15:24	`DM556`	Explainable Skill Acquisition over Time via GraphRAG-Augmented Mastery Features, Fuzzy Clustering, and Hybrid Deep Models	Qiong Cheng

Oral Presentations · 16:00–17:30

← Back to agenda

UH7 — Federal A UH8 — Federal B UH9 — South American B UH10 — California UH11 — New York UH12 — Massachusetts

16:00–17:30 · Oral Presentations（12 mins × 7 papers × 6 rooms）

Session Chairs — UH7–UH12

UH7

Yi He

UH8

Kunpeng Liu

UH9

Kanthi K Sarpatwar

UH10

Zhou Yang

UH11

Senjuti Basu Roy

UH12

Meikang Qiu

UH7 (Room: Federal A, Session Chair: Yi He): S01284, S01283, S01240, S01239, S01238, S01237, S01236

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`S01284`	Multi-optimizer Deep&Cross at Industrial Scale	Mark znidar
2	16:12–16:24	`S01283`	Multi-Modal Embedding Fusion for Scalable Context-First CTR	Mark znidar
3	16:24–16:36	`S01240`	A Deep Learning Approach for Reaction-Diffusion-Advection Modeling of Vegetation-Desertification Patterns	Aritro Chatterjee
4	16:36–16:48	`S01239`	Dynamics of Fencer Rating Progression	Ethan Xu
5	16:48–17:00	`S01238`	ESN-DAGMM: A Lightweight Framework for Unsupervised Time-Series Data Monitoring in 5G O-RAN Networks	Raymond Zhao
6	17:00–17:12	`S01237`	Deep Gaussian Fusion Network for Traffic Prediction	Zhiqian Chen
7	17:12–17:24	`S01236`	Quantifying Biopharma Alliance Fragility Using a Strategic Shock Risk Index (SSRI)	Rhea Zhou

UH8 (Room: Federal B, Session Chair: Kunpeng Liu): S01235, S01234, S01233, S01232, S01231, S01230, S01229

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`S01235`	Do You Know What I Mean? Testing the Prompt Robustness of an LLM-Powered IoT System	Xingguo Ding
2	16:12–16:24	`S01234`	Interactive 3D Spine Modeling for Enhanced Doctor-Patient Communication and Health Literacy	Christian Jin
3	16:24–16:36	`S01233`	AI or humans: Who understands online emotions Better?	Victor Tang
4	16:36–16:48	`S01232`	Mining Mobile Point-of-Interest Visit Data for Socioeconomic Insights	Amy Ma
5	16:48–17:00	`S01231`	An Agentic Framework for Social Event Forecasting: Approaches using Causality Contextualized Chain of Thought	Avani Thakur
6	17:00–17:12	`S01230`	Data-Driven Weakly-Supervised Methods Successfully Denoise Diverse Biomedical Imaging Modalities	Reeti Rout
7	17:12–17:24	`S01229`	Lost in Transcription: Influence of Dialect on Whisper’s Performance	Helen Qin

UH9 (Room: South American B, Session Chair: Kanthi K Sarpatwar): DM314, S01228, S01227, S01226, S01224, S01223, S01222

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`DM314`	Benchmarking LLMs and Distributed Approaches for Anomaly Detection	Qiong Cheng
2	16:12–16:24	`S01228`	Grounded Chest X-Ray Reasoning: Leveraging Visual Tools to Improve Medical Multimodal LLMs	Huaxiu Yao
3	16:24–16:36	`S01227`	Bringing Optimization to Everyone: Exploring LLMs as a Tool for Non-Experts and Students	Winston Zhang
4	16:36–16:48	`S01226`	Deep Learning to Denoise and Segment Air Pollutant Plumes	Daniel Li
5	16:48–17:00	`S01224`	EEG EyeNet: Strong, Insightful Baselines for Eye-Movement Prediction from EEG	Christian Jin
6	17:00–17:12	`S01223`	Automated Analysis of Astrocyte Cell Connectivity after Laser Ablation Using Machine Learning and Path-Finding Algorithms	Connor Lee
7	17:12–17:24	`S01222`	Predicting Vertical Cloud Type Structure with GOES-ABI Multi-Channel Data (Deep Learning & Foundation Model)	Sidh Jaddu

UH10 (Room: California, Session Chair: Zhou Yang): S01221, S01220, S01219, S01218, S01217, S01216

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`S01221`	Multimodal Foundation Models as Router Models for High-Resolution Aerial Image Segmentation	Cooper Li
2	16:12–16:24	`S01220`	Forecasting U.S. Recessions with Machine Learning: Evidence from Ten Economic Indicators, 1978–2025	Neel Dhuruva
3	16:24–16:36	`S01219`	C-Reactive Protein Induces Endothelial Cell Dysfunction and Replication Stress	Jay Peng
4	16:36–16:48	`S01218`	Systematic Comparison of Artifact Removal Techniques for Reliable Feature Extraction from scTS-Contaminated EMG	Vivian Li
5	16:48–17:00	`S01217`	Graph-LLM for EHRs: Combining Temporal Graph Representations and LLM-Based Note Imputation for Clinical Predictions	Michael Liu
6	17:00–17:12	`S01216`	LLMSeqRec: LLM Enhanced Contextual Sequential Recommender	Connor Lee

UH11 (Room: New York, Session Chair: Senjuti Basu Roy): S01215, S01214, S01213, S01212, S01211, S01210

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`S01215`	Benchmarking the Code Generation Capabilities of Popular Large Language Models for Front-End Web Development	Dron Datta
2	16:12–16:24	`S01214`	Persona-Driven LLM Interaction in Stock Market Simulations	Medhashree Parhy
3	16:24–16:36	`S01213`	Evaluating the Effectiveness of Persona Simulation in Opinion Prediction with GPT-4.1	Sarah Li
4	16:36–16:48	`S01212`	AI-Powered Trait Analysis for Poisonous Mushroom Classification	Srikar Akundi
5	16:48–17:00	`S01211`	Machine Learning-Based Classification of Transcriptional Gene Groups for Cancer Prognosis	Annie Wu
6	17:00–17:12	`S01210`	Performance Evaluation of Convolutional Neural Networks in Image-Based Malware Classification	Raymond Jiang

UH12 (Room: Massachusetts, Session Chair: Meikang Qiu): S01208, S01207, S01206, S01204, S01203, S01201

#	Time	Paper ID	Title	Contact
1	16:00–16:12	`S01208`	Signature vs. Substance: Evaluating the Balance of Adversarial Resistance and Linguistic Quality in Watermarking Large Language Models	William Guo
2	16:12–16:24	`S01207`	Hyperspectral Band Selection with Learnable Weights for Efficient Glioblastoma Detection	Albert Li
3	16:24–16:36	`S01206`	Dimension Reduction Enhanced Boosting for Imbalanced Data Classification	Eric Wang
4	16:36–16:48	`S01204`	Probabilistic Prompts for Zero-shot and Few-shot Large Language Models: An Empirical Study of Patient-reported Outcomes	Zhong Chen
5	16:48–17:00	`S01203`	Quantitative Assessment on the Impact of Music on Athletic Performance	Angela Du
6	17:00–17:12	`S01201`	Predicting Relationship Stability Using Communication Patterns	Deepak Gahalot

Parent Panel — Panelists

← Back to agenda

This panel appears on Nov 15, 11:30–12:25 (see agenda above).

Parent Panelists
Name	Email	Affiliation	Title / Role
Min Wu	minwu@umd.edu	University of Maryland – College Park	Distinguished University Professor; Associate Dean for Graduate Programs; Christine Yurie Kim Eminent Professor in Information Technology, Electrical and Computer Engineering
Ambreen Hasan	ambihasan@gmail.com	Youngstown State University	Director of Institutional Research and Analytics
Jianjun Xie	jianjunxie@gmail.com	FICO	Principal Scientist
Dhuruva Badri	drubadri@gmail.com	Florida Department of Transportation	Assistant District Construction Engineer
Tarun Thakur	tarun.thakur@alumni.duke.edu	Veza	Co-Founder and CEO

Poster Presentation

The recommended poster size is A0 (33.1″ × 46.8″ or 841 mm × 1189 mm). Posters may be prepared in either portrait or landscape orientation. Authors are free to design their posters as they see fit, as long as the poster fits within the allotted display area.

Each poster board measures 8 feet wide by 4 feet high and is double-sided. Two posters will be displayed on each side, so each presenter has a maximum space of 4 feet × 4 feet (1.22 m × 1.22 m), including margins. For visual reference, please see the attached diagram illustrating the poster board layout.

8’ board width

Illustration: one 8’ × 4’ board (one side) with two 4’ × 4’ poster areas.
Presentation materials may be attached using pushpins or tape, which will be provided on site.

Scope of Topics

We invite submissions of original research papers from undergraduate and high school students on topics related to data mining, including but not limited to:

Topics of Interest

Foundations, algorithms, models, and theory of data mining, including big data mining
Machine learning, deep learning, and statistical methods for big data
Mining heterogeneous data sources, including text, semi-structured, spatio-temporal, streaming, graph, web, and multimedia data
Data mining systems and platforms for analyzing big data, including methods for parallel and distributed data mining, federated learning, and their efficiency, scalability, security, and privacy
Data mining for modeling, visualization, personalization, and recommendation
Data mining for cyber-physical systems and complex, time-evolving networks
Data mining with large language models
Novel applications of data mining in data science, including big data analysis in social sciences, physical sciences, engineering, life sciences, climate science, web, marketing, finance, precision medicine, health informatics, and other domains

Submitted papers should present novel ideas, methodologies, algorithms, or applications in the realm of data mining. Papers will be evaluated based on their technical quality, novelty, relevance, and clarity of presentation.

Eligibility

Undergraduate students and high school students pursuing an academic degree at the time of submission are eligible to submit papers as first authors.
Each submission must have at least one student author, who should be the presenter if the paper is accepted.
Co-authorship with faculty members or researchers is allowed, but the student must be the primary contributor to the work.

In-Person Policy (UGHS)

The UGHS Symposium will be run as an in-person session. At least one student author must attend and present the paper in person; otherwise, the paper will not be included in the IEEE proceedings.

Additional requirements for high school student authors:

At least one parent accompanies the student to the conference.
A signed release form (download here).
An information document signed by a parent/guardian, along with an optional signature from the high school principal. The document should include the paper ID and title; contact information for the parent (required) and the principal (optional); a statement of permission from the high school (optional); and confirmation that a parent will accompany the student.
Template: Information Document (DOCX)

Note: School approvals are optional. However, some schools require their permission.

The conference policy requires that each student participant be accompanied by their own parent or legal guardian. One adult chaperone cannot supervise multiple students.

Where to send: Please email the signed release form and the optional information document to ieee-icdm-2025-undergraduate-and-high-school-symposium-g@vt.edu before November 7, 2025.

Submission Format Requirements

Submissions must adhere to the IEEE Computer Society Proceedings Manuscript Formatting Guidelines (see link to “formatting instructions” below: https://www.ieee.org/conferences/publishing/templates.html).
Undergraduate student research papers should not exceed 6 pages, including all figures, tables, and references.
High school student research papers should not exceed 5 pages, including all figures, tables, and references.
Please highlight whether the first author is a high school or an undergraduate student in the author affiliation of your submitted paper.

Publication Ethics and Dual Submission Policy

We uphold the highest standards of academic integrity and ethical research conduct. All authors must ensure that their submissions fully comply with the following guidelines.

Originality and Dual Submission: Submitted manuscripts must represent original work that has not been submitted or published elsewhere. Dual submission—submitting the same or substantially similar content to multiple venues concurrently— is strictly prohibited and constitutes a violation of publication ethics and integrity.

Subsequent Journal Submission: If authors wish to submit an extended version of their published work to a journal after conference publication (e.g., IEEE ICDM proceedings), they must clearly disclose that a preliminary version has been published. Transparency with journal editors and reviewers is required. Any subsequent submission to another venue must include significant new contributions, such as expanded experiments, novel theoretical insights, or additional methodological developments. Minor revisions or incremental changes are not sufficient to qualify as a new work.

Ethical Responsibility: Authors are responsible for maintaining ethical standards in all aspects of the publication process, including authorship, data integrity, and proper citation of prior work. Violations may lead to retraction, notification of the authors’ institutions, and disqualification from future submissions.

Important Dates

Paper Submission Deadline: ~~Sep 1, 2025 11:59 PM AoE~~
Paper Submission Deadline (Deadline Extended): ~~Sep 10, 2025 11:59 PM AoE~~
Notification of Acceptance: ~~Oct 1, 2025~~
Camera-Ready Paper Submission: ~~Oct 15, 2025~~
Symposium Date: Nov 15, 2025

Venue & Hotels

IEEE ICDM 2025 will be held in-person in Washington, D.C. Discounted hotel rooms are available. Please check: https://www3.cs.stonybrook.edu/~icdm2025/venue.html .

How to Submit

Open Submission Portal Camera-Ready Instructions

The review process is single-blind (reviewers anonymous; authors visible to reviewers).
Accepted papers will appear in the ICDM Workshop Proceedings (IEEE Computer Society Press) and be available at the conference.
Each accepted paper must have at least one author registered and present in person during the symposium day.

Camera-Ready Submission

Final deadline: October 15, 2025.

The camera-ready paper must follow all formatting requirements (not a response to reviewer comments). Ensure IEEE formatting/page limits per the ICDM 2025 website, use high-clarity figures, and keep the layout clean and consistent.
Submission site: IEEE CPS Login
PDF eXpress validation: Conference Record 69685.
Paper ID on the registration form: use your conference submission ID (e.g., S01217) from the submission emails — not the IEEE PDF eXpress ID (e.g., 2025357692).
Issues with copyright or submission? Contact Martha Nunez.

Registration

We have already provided the discount registration code in the camera-ready email to all authors. That discounted registration fee is the one-day symposium fee, rather than the full conference fee. It covers one student and one parent companion, and also includes the Friday banquet.

Program Co-Chairs

Join us in shaping the future of data mining research by sharing your insights and discoveries at the Undergraduate and High School Symposium. For questions, please contact the symposium co-chairs.