Hi! I’m currently a Ph.D. candidate at Worcester Polytechnic Institute in our Data Science program, a part of the Computer Science department. Broadly, I am interested in sequence modeling and so my research has revolved heavily around Recurrent Neural Networks. Ultimately I aim to advance sustainability in machine learning and apply machine learning to sustainability challenges, motivated by the low power consumption of the human brain and the massive carbon footprint of modern deep learning.
Specific Research Interests
I am interested in sequence modeling, or building vector representations that capture relevant temporal dynamics in sequential data such as time series or text. So far, I have studied early classification, clinical note classification, meta word embeddings, explainable neural time series classification, and representation learning for irregularly-sampled time series. Most of my work involves some form of conditional computing in RNNs. I have also had the pleasure of advising many NSF-funded REU students over the summers on some research involving missing values in clinical time series and sequential diagnosis prediction using RNNs.
Please feel free to contact me with questions regarding my research or our program at WPI at twhartvigsen ‘at’ wpi ‘dot’ edu. I would be happy to go into much more detail on my research one-on-one.
- [HEALTHINF] Erin Teeple, Thomas Hartvigsen, Cansu Sen, Kajal Claypool, Elke Rundensteiner. Clinical Performance Evaluation of a Machine Learning System for Predicting Hospital-Acquired Clostridium Difficile Infection. HEALTHINF, 2020. [Paper].
- [BigData] Cansu Sen, Thomas Hartvigsen, Xiangnan Kong, Elke Rundensteiner. Patient-Level Classification of Clinical Note Sequences Guided by Attributed Hierarchical Attention. IEEE BigData, 2019. [IEEE Link].
- [BigData] Cansu Sen, Thomas Hartvigsen, Xiangnan Kong, Elke Rundensteiner. Learning Temporal Relevance in Longitudinal Medical Notes. IEEE BigData, 2019.[IEEE Link].
- [KDD] Thomas Hartvigsen, Cansu Sen, Xiangnan Kong, Elke Rundensteiner. Adaptive-Halting Policy Network for Early Classification. ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2019. [ACM Paper Link][pdf][code].
- [BHI] Jidapa Thadajarassiri, Cansu Sen, Thomas Hartvigsen, Xiangnan Kong, Elke Rundensteiner. Comparing General and Locally-Learned Word Embeddings for Clinical Text Mining. IEEE International Conference on Biomedical and Health Informatics (BHI), 2019. [pdf].
- [CCIS] Thomas Hartvigsen, Cansu Sen, Elke Rundensteiner. Detecting MRSA Infections by Fusing Structured and Unstructured Electronic Health Record Data. Communications in Computer and Information Science. [pdf]
- [HEALTHINF] Thomas Hartvigsen, Cansu Sen, Sarah Brownell, Erin Teeple, Xiangnan Kong, Elke Rundensteiner. Early Prediction of MRSA Infections using Electronic Health Records. International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2018) - Volume 5: HEALTHINF, pages 156-167, ISBN: 978-989-758-281-3. Shortlisted for Best Student Paper. [pdf]
- [ECML/PKDD] Cansu Sen, Thomas Hartvigsen, Kajal Claypool, Elke Rundensteiner. CREST - Risk Prediction for Clostridium Difficile Infection Using Multimodal Data Mining. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECMLPKDD) 2017. [pdf]