Software Engineer Developer
Resume:
ROBBY J. M. GOETSCHALCKX
Present Address
OR 97401 Eugene
Oregon
USA
Permanent Address
Smeyerspad 7
Belgium
*****.************@*****.***
Citizenship
Belgium
US permanent resident
Employment
December 2017 { June 2020: Connected Signals:
Senior Software Engineer
Building predictive models to predict tra c lights, bus arrival times at signalized inter- sections, and other tra c-related predictions.
Data analysis, machine learning, data visualization.
Languages: Python, C++.
Extensive experience with Pandas DataFrames.
September 2017 { December 2017: Independent contractor: Senior Software Engineer / Researcher
Continuing collaboration with Eduworks Corporation. January 2016 { September 2017: Eduworks Corporation: Software Developer / Researcher
Working on Machine Learning and Natural Language Processing problems.
Projects include automated question generation from text documents, inference in com- petency frameworks (predicting con dence levels that a particular person does or does not hold a given competency based on assertions of other competencies), detection of voice phishing attacks.
Languages: Java, Python as well as an in-house functional language. December 2012 { December 2015: Oregon State University: Post-Doctoral Researcher
Supervisors: Prasad Tadepalli, Alan Fern
Working on the Active Transfer Learning project, assisting in supervision of graduate students. Focus on multi-agent planning, hierarchical policy learning, multi-task learning, active learning, active tutoring.
Sept. 2011 { Aug. 2012: University of Waterloo / University of Toronto: Post-doctoral Fellow
Supervisors: Jesse Hoey (UW), Pascal Poupart (UW), Alex Mihailidis (UofT)
Continuation of the work on Machine Learning for autonomous post-stroke rehabilita- tion from my position in Dundee, now working closer together with the people from the University of Toronto and the Toronto Rehab Institute. Fundamental research in Bayesian Machine Learning for sequential decision problems is combined with research into practical and feasible solutions for the speci c eld of stroke rehabilitation. Feb. 2010 { June 2011: University of Dundee:
Post-doctoral Research Assistant
Supervisors: Jesse Hoey, Stephen McKenna
DyNaMo Project, an EU / Mexican government project, FONCICYT 95185
The project involves the development of a haptic-robotic device used in stroke rehabil- itation. In the project we researched and developed a POMDP model for the system, allowing a policy which balances information gathering on the characteristics of the per- son who su ered a stroke with quick and successful rehabilitation. Education
Katholieke Universiteit Leuven (Belgium):
Ph.D. Computer Science, 2009
thesis title: The Use of Domain Knowledge in Reinforcement Learning Advisors: Hendrik Blockeel, Maurice Bruynooghe
Specialization: Machine Learning, Reinforcement Learning Advanced Studies: Master of Arti cial Intelligence, 2004 Cum Laude
thesis title: Phase transition in Answer set Programming Supervisors: Marc Denecker, Maarten Mari en
Licentiate Informatica (decreed equivalent to Master’s degree in informatics), 2003 Cum Laude
Specialization: AI, theoretical computer science
Thesis title: The Smallest Universal Turing Machine Supervisor: Bart Demoen
Patents
US Patent #10,614,325 \Automatic detection of tra c signal states for driver safety system" US Patent #10,614,106 \Automated tool for question generation" Languages
Dutch: Native language
English: Native language pro ciency
French: Limited working pro ciency (8 years)
German: Passive: good; Active: notions (2 years)
Skills
Research and development of Machine Learning algorithms and methods Research and development of Sequential Decision Making / Planning algorithms and methods Data science
Advanced mathematics
Programming (Java, Python, Prolog)
Scienti c and technical writing
Teaching, tutoring and mentoring
Publications
Selected list of publications:
Goetschalckx, R., Fern, A., Tadepalli, P.
Multitask Coactive Learning.
Proceedings of the Twenty-Fourth International Joint Conference on Arti cial Intelligence
(IJCAI). 2015.
Hamidi, M., Tadepalli, P., Goetschalckx, R., Fern, A. Active Imitation Learning of Hierarchical Policies. Proceedings of the Twenty-Fourth International Joint Conference on Arti cial Intelligence
(IJCAI). 2015.
Goetschalckx, R., Fern, A., Tadepalli, P.
Coactive Learning for Locally Optimal Problem Solving. Proceedings of the Twenty-Eighth AAAI Conference (AAAI). (2014). Judah, K., Farn, A., Tadepalli, P., Goetschalckx, R. Imitation Learning with Demonstrations and Shaping Rewards. Proceedings of the Twenty-Eighth AAAI Conference (AAAI). (2014). Goetschalckx, R., Poupart, P., Hoey,J.
Continuous Correlated Beta Processes.
Proceedings of the Twenty-second International Joint Conference on Arti cial Intelligence
(IJCAI). (2011).
Kan, P., Huq, R., Hoey, J., Goetschalckx, R.,Mihailidis, A. The Development of an Adaptive Upper-limb Stroke Rehabilitation Robotic System. Journal of NeuroEngineering and Rehabilitation. (2011). Huq, R., Kan, P., Goetschalckx, R., H ebert, D., Hoey, J., Mihailidis, A. A Decision-Theoretic Approach in the Design of an Adaptive Upper-Limb Stroke Rehabilitation Robot.
Proceedings of the 12th International Conference on Rehabilitation Robotics. (2011). Goetschalckx, R., Driessens, K.
Parsimonious Linear Model Trees.
Proceedings of the ICML 2010 Workshop on Machine Learning and Games. (2010). R. Goetschalckx, O. Missura, J. Hoey, T. Gaertner. Games with Dynamic Di culty Adjustment through the Use of POMDPs. Proceedings of the ICML 2010 Workshop on Machine Learning and Games. (2010). Sanner, S., Goetschalckx, R., Driessens, K., and Shani, G. Bayesian Real-time Dynamic Programming.
Proceedings of the 21st International Joint Conference on AI (IJCAI-09). (2009). Goetschalckx, R., Sanner, S., Driessens, K.
Reinforcement Learning with the Use of Costly Features. Post-Proceedings of the European Workshop on Reinforcement Learning (EWRL 08), Lecture notes in computer science, 5323/2008, art.nr. 10, (pp. 124{135). (2008). Goetschalckx, R., Sanner, S., Driessens, K.
Cost-sensitive Parsimonious Linear Regression.
Proceedings of the 8th IEEE International Conference on Data Mining. ICDM. Pisa, Italy
(pp. 809{814). IEEE Computer Society. (2008).
Goetschalckx, R., Driessens, K.
Cost-sensitive Reinforcement Learning.
Proceedings of the workshop on AI Planning and Learning. Providence, Rhode Island, USA
(pp. 1{5). (2007).
Goetschalckx, R., Ramon, J.
Using Expert Knowledge to Construct Error Bound State-action Aggregations for Reinforce- ment Learning.
Proceedings of the 19th Belgian-Dutch Conference on Arti cial Intelligence. 19th Belgian- Dutch Conference on Arti cial Intelligence. Utrecht, The Netherlands (pp. 127{134). (2007). Goetschalckx, R., Ramon, J.
On Policy Learning in Restricted Policy Spaces.
AAAI 2007 Student Abstract and Poster Program., Vancouver, Canada. (2007). Tuyls, K., Croonenborghs, T., Ramon, J., Goetschalckx, R., Bruynooghe, M. Multi-agent Relational Reinforcement Learning.
Proceedings of the First International Workshop on Learning and Adaptation in Multi-Agent Systems. First international workshop on Learning and Adaptation in Multi Agent Systems held at the 4th International Conference on Autonomous Agents and Multi-Agents Systems. Utrecht, The Netherlands, (pp. 123{132). (2005).
Teaching
Oregon State University: Occasionally lling in for professor Prasad Tadepalli, \Arti cial Intelligence" 2013-2014.
Teaching Assistant for the course \Fundamentals of Arti cial Intelligence" 2004{2009, in the programmes: ‘Bachelor in Informatics’, ‘Engineering: Computer Science’ and ‘Advanced stud- ies: Master of Arti cial Intelligence’.
Students Supervised
Graduate students supervised (Oregon State University): Kranti Kumar Potanappali (MS 2014): Learning for Search and Coverage Mandana Hamidi: Imitation Learning of Hierarchical Policies B eatrice Moissinac: Intelligent Tutoring
Kshitij Judah: (PhD 2014) New Learning Modes for Sequential Decision Making Master theses supervised (K.U.Leuven):
Ewoud Nuyts: General Game Learning, 2008{2009
Joaquin Vanschoren: Development of a framework for high-level perception, 2004{2005 Other Responsibilities
Ombudsman for the Advanced studies: Master of Arti cial Intelligence’ programme: 2005{ 2008
This job entails mediating in case of con
ict between students and teachers, as well as advising students in case of irregularities during exam period. Representative of teaching assistants in the programme advisory committee for the ‘Advanced studies: Master of Arti cial Intelligence’ programme: 2005{2008 Student representative in the programme advisory committee for the ‘Advanced studies: Mas- ter of Arti cial Intelligence’ programme: 2003{2004
Contact this candidate