Ali el Hassouni

PhD candidate in Artificial Intelligence

Biography

I am a Ph.D. candidate in the Computational Intelligence group at the Vrije Universiteit Amsterdam (VU) and in Mobiquity’s Data Science and Analytics team. I’m interested in data efficient (Deep) Reinforcement Learning (RL) applied to real-world problems. To achieve this I’ve developed a simulator for daily human behavior/activities powered by generative models (GANs) and recurrent neural networks (LSTMs). The simulator is used as a testbed for novel data efficient RL algorithms. The ultimate aim is transfer learning from simulation to real-world.

My Ph.D. project aims to develop data efficient RL algorithms that can underpin mobile applications and strive to improve the level of personalisation in the health domain. The main objective is to develop methods that allow the app to provide interventions to users in an intelligent and proactive manner by learning to select suitable actions for particular situations. My project supervisors are Gusz Eiben (VU), Mark Hoogendoorn (VU) and Vesa Muhonen (Mobiquity).

Interests

Artificial Intelligence
Machine Learning
Reinforcement Learning
Deep Learning
Simulation
Personalisation
Healthcare

Education

MSc in Business Analytics, Cum Laude, 2016

Vrije Universiteit Amsterdam
BSc in Business Analytics, 2014

Vrije Universiteit Amsterdam

Recent Publications

Quickly discover relevant content by filtering publications.

End-to-end Personalization of Health Interventions Using Raw Sensor Data With Deep Reinforcement Learning (In submission)

Ali el Hassouni, Mark Hoogendoorn, A. E. Eiben, Martijn van Otterlo, Vesa Muhonen

Using Generative Adversarial Networks to Develop a Realistic Human Behavior Simulator

Simulation environments have proven to be very useful as testbeds for reinforcement learning (RL) algorithms. For settings where an …

Ali el Hassouni, Mark Hoogendoorn, Vesa Muhonen

PDF Poster Slides DOI

Personalization of Health Interventions Using Cluster Based Reinforcement Learning

Research has shown that personalization of health interventions can contribute to an improved effectiveness. Reinforcement learning …

Ali el Hassouni, Mark Hoogendoorn, Martijn van Otterlo, Eduardo Barbaro

PDF Poster Slides DOI

Prediction Using Patient Comparison vs Modeling: A Case Study for Mortality Prediction

Information in Electronic Medical Records (EMRs) can be used to generate accurate predictions for the occurrence of a variety of health …

Mark Hoogendoorn, Ali el Hassouni, Kwongyen Mok, Marzyeh Ghassemi, Peter Szolovits

Preprint PDF Code DOI

Student supervision

Performance of exploration methods using DQN - Tim Elfrink - July 2018
Master thesis Business Analytics - Vrije Universiteit Amsterdam and Mobiquity Inc

Image Verification with Siamese Capsule Networks - Mark de Blaauw - Jan 2019
Research paper Business Analytics - Vrije Universiteit Amsterdam

Generate sensory data using CGAN - Maarten Kleinrensink - Feb 2019
Research paper Business Analytics - Vrije Universiteit Amsterdam

Generating mood patterns using CGANs - Steven Postmus - March 2019
Master thesis Business Analytics - Vrije Universiteit Amsterdam (Co-supervised with Mark Hoogendoorn)

Improving Deep Clustering via Adversarially Constrained Interpolation - Daniel Lutscher - Ongoing
Master thesis Artificial Intelligence - Vrije Universiteit Amsterdam and Braincreators

Using CNNs for time series prediction - Mihail Tomescu - Ongoing
Master thesis Artificial Intelligence - Vrije Universiteit Amsterdam

Generating raw sensor data with Conditional Generative Adversarial Networks - Madelyn Deuten - Ongoing
Bachelor Computer Science - Vrije Universiteit Amsterdam

Prediction of churn and Customer Lifetime Value on a subscription-based service using behavioural and transactional data
Daniela Zamora Uribe - Ongoing
Master thesis Business Analytics - Vrije Universiteit Amsterdam and bunq

Teaching Assistant

Heuristics - Bsc. course - 2015-2016 - Vrije Universiteit Amsterdam (VU)

Machine Learning - Msc. course - 2015-2016 - Vrije Universiteit Amsterdam (VU)

Computational Intelligence - Msc. course - 2015 - Universiteit van Amsterdam (UVA)

Machine Learning for the Quantified Self - Msc. course - 2016-2017 and 2017-2018 - Vrije Universiteit Amsterdam (VU)

Contact

Room T-316, de Boelelaan 1081a, 1081HV Amsterdam, Netherlands
Skype Me