
My name is Floris den Hengst and I am a PhD. student in the Quantitative Data Analytics and Knowledge Representation & Reasoning groups of the Vrije Universiteit Amsterdam and ING’s AI for FinTech Research. I am interested in adaptivity v.s. control in learning systems, knowledge representation and language systems.
My PhD research is targeted at making Reinforcement Learning agents comply to regulations. Adaptivity can be of huge benefit in highly regulated domains such as healthcare and finance. In these domains, systems have to comply to regulations before they can be used. Understanding the regulations, on the other hand, is an expertise in itself.
This project touches practical and fundamental aspects of explainability and safety of AI. How can we formalize regulations so that domain experts can inspect, understand and validate the result? How can we bridge the gap between experts' and RL agents' representation of the world and actions? How does constraining a RL agent impact its learning capabilities? In this project, we use an adaptive conversational agent for financial advice to investigate these issues.
Experience
PhD student, Vrije Universiteit Amsterdam and ING, current
Software Engineer, Crunchr, 2014 - 2017
Msc. Artificial Intelligence, Cum Laude, Vrije Universiteit Amsterdam, 2013 - 2016
Web Developer, Zilt&Co, 2009 - 2014
Various teaching- and research assistant positions, 2009 - 2014
Bsc. Artificial Intelligence, GPA 7.9 / 10, Vrije Universiteit Amsterdam, 2008 - 2013
Officially titled `Lifestyle Informatics' at the time
Publications
Smit, Yannick and den Hengst, Floris and Bhulai, Sandjai and Mehdad, Ehsan,
“Strategic Workforce Planning with Deep Reinforcement Learning”.
to appear at LOD (2022)
pdf
Den Hengst, Floris and François-Lavet, Vincent and Hoogendoorn, Mark and Van Harmelen, Frank,
“Reinforcement Learning with Option Machines”.
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, 2909-2915 (2022)
doi pdf preprint bib
Den Hengst, Floris and François-Lavet, Vincent and Hoogendoorn, Mark and Van Harmelen, Frank,
“Planning for potential: efficient safe reinforcement learning”.
Machine Learning, Springer (2022)
Presented at BeNeRL 2022
doi bib pdf
Den Hengst, Floris and Grua, Eoin Martino and el Hassouni, Ali and Hoogendoorn, Mark,
“Reinforcement Learning for Personalization: A Systematic Literature Review”.
Data Science (2020)
Presented at RL for Real Life conference 2020, presented at BNAIC 2020
doi bib
Van Zeelt, Mickey and Den Hengst, Floris and Hashemi, Seyyed Hadi,
“Collecting High Quality Dialogue User Satisfaction Ratings with Third-Party Annotators.”
Proceedings of the 2020 Conference on Human Information Interaction and Retrieval,
363-367 (2020)
doi bib
Den Hengst, Floris and Hoogendoorn, Mark and Van Harmelen, Frank and Bosman, Joost,
“Reinforcement Learning for Personalized Dialogue Management”. 2019 IEEE/WIC/ACM International
Conference on Web Intelligence (WI), 59-76 (2019)
doi bib pdf
Floris den Hengst - Detecting Interesting Outliers: Active Learning for Anomaly Detection,
Master’s thesis Artificial Intelligence, Vrije Universiteit Amsterdam (2016)
pdf
Student supervision
Current:
- Haritha Jayaraman, Msc. thesis
Past:
- Stefan Petrescu, “Rethinking Log Parsing in the Context of Modern Software Ecosystems”, Msc. thesis
- Daniel van der Riet, “Safe Reinforcement Learning with a Learned Transition Function”, Msc. thesis
- Yannick Smit, “Strategic Workforce Planning with Deep Reinforcement Learning”, Msc. thesis
- Claudia-Violeta Grigorias, “Learning safe behaviours with dynamic hyperparameters in Reinforcement Learning”, Bsc. thesis
- Michal Nauman, “Low-Variance Policy Gradient Estimation with World Models”, Msc. thesis
- Azamat Omuraliev, “Reinforcement Learning for Controllable Text Summarization”, Msc. thesis
- Mickey van Zeelt, “Gathering External Evaluations on Chatbot Conversations”, Msc. thesis
- Tim Nederveen, “Reducing Expert Interference in Time Series Anomaly Detection Model Re-evaluations”, Msc. thesis
- Rob Wanders “Predicting Number of Transactions with Echo State Networks”, BA paper
- Luca Simonetto “Generating Spiking Time Series with Generative Adversarial Networks: an Application on Banking Transactions”, Msc. thesis
- Ilse Goedhart “Predicting IT Performance using Quantile Regression”, BA paper