Machine Learning Researcher @ Sony | Oct 2023 - now

Project Lead & Research Scientist | Jul 2021 - Jun 2023

• Built a neighbor relation graph learning framework for real estate appraisal at E.SUN bank to reduce the appraiser’s workload by about 80%.

• Applied explainable techniques to analyze the model behavior which helps the appraiser understand the model.

• Led a 3-member team, cooperated with the engineering team, and helped deploy the model to production.

Machine Learning Scientist Intern @ Appier | Jun 2022 - Nov 2022

• Implemented an ensemble real-time bidding model with new data-driven features from 10M+ e-commerce click stream data.

• Conducted tree analysis and feature importance with SHAP to analyze model behavior.

• Cooperated with 3 data scientists to build the RTB model for re-engagement campaigns.

Teaching Assistant of Introduction to Artificial Intelligence, NYCU | Feb 2022 - Jun 2022

• Designed 2 real-world AI assignments including using yolov5 to detect the parking space and adopting LIME and SHAP to explore the pre-trained language model behavior.

• Gave a keynote about the progress of Explainable AI to 100+ students.

Data Scientist Intern @ China Engineering Consultants | Jul 2020 - Dec 2020        

• Built a time series model to predict pavement damage with various domain experts which was used by New Taipei City Government.

• 3rd place in Digi+ talent accelerator & jumpstart competition (total 40+ teams)

Co-founder, Leafhopper (a data-driven smart agriculture startup team) | Mar 2019 - Jun 2020

• Member of NTHU Garage and 3rd place in the 2020 NTHU Entrepreneur Day.

• Attracted over 20,000+ people’s attention to our project.

Chih-Chia Li, Wei-Yao Wang, Wei-Wei Du, Wen-Chih Peng, ”Look Around! A Neighbor Relation Graph Learning Framework for Real Estate Appraisal”, PAKDD 2024 [Paper]

Wei-Wei Du, Wei-Yao Wang, Wen-Chih Peng, ”DoRA: Domain-Based Self-Supervised Learning Framework for Low-Resource Real Estate Appraisal”, CIKM 2023 [First Author] [Paper] [Code]

Wei-Yao Wang, Wei-Wei Du, Wen-Chih Peng, ”ShuttleSet22: Benchmarking Stroke Forecasting with Stroke-Level Badminton Dataset”, IJCAI 2023 IT4PSS Workshop [Second Author] [Paper] [Code]

Wei-Wei Du, Hong-Wei Wu, Wei-Yao Wang, Wen-Chih Peng, ”Team Triple-Check at Factify 2: Parameter-Efficient Large Foundation Models with Feature Representations for Multi-Modal Fact Verification”, AAAI 2023 Workshop on Multimodal Fact Checking and Hate Speech Detection [First Author] [Paper] [Code]

Chih-Chia Li, Wei-Yao Wang, Wei-Wei Du, Wen-Chih Peng, ”Look Around! A Neighbor Relation Graph Learning Framework for Real Estate Appraisal”, AAAI 2023 MuFin Workshop [Best Paper Award] [Third Author] [Paper]

Wei-Wei Du, Wei-Yao Wang, Wen-Chih Peng, ”Track2Vec: Fairness Music Recommendation with a GPU-Free Customizable-Driven Framework”, CIKM 2022 EvalRS Workshop [First Author] [Paper] [Medium] [Code]

Wei-Yao Wang, Wei-Wei Du, Wen-Chih Peng, ”RecFormer: Personalized Temporal-Aware Transformer for Fair Music Recommendation”, CIKM 2022 EvalRS Workshop [Second Author] [Paper] [Code]

Wei-Wei Du*, Wei-Yao Wang*, Yu-Chien Tang*, Wen-Chih Peng, ”NYCU_TWD@LT-EDI-ACL2022: Ensemble Models with VADER and Contrastive Learning for Detecting Signs of Depression from Social Media”, ACL 2022 LT-EDI Workshop (*indicates equal contribution) [First Author] [Paper] [Medium] [Code]