Responsible for the large scale integration of automated vector autogressive modeling and analysis for several research projects and websites that use EMA (Ecological Momentary Assessments) data to assess well-being. Experienced full-stack development as part of a multidisciplinary team for two of these projects.• Hoe gek is Nederland? (translation: "How nuts are the Dutch?") is an online platform designed to support self-measurement of mental health for the entire population of the Netherlands by allowing them to fill out questionnaires and by providing them with personalized feedback. Currently, the web application has ~12,000 active users. Personalized feedback is provided by applying Autovar to find VAR models that determinine contemporary and causal relations between quality of life indicators of EMA data. Those relations are then visualized as graphs to provide insight to the user, e.g., for identifying the indicators that have a high likelihood of being significant for improving their well-being. To the best of my knowledge, this is the first fully-automated derivation and application of vector autoregression models.• Leefplezier (translation: "Well-being") is an online platform to support self-measurement of mental health for the elderly population of the Netherlands by allowing them to fill out questionnaires and by providing them with personalized feedback. The aim is to provide the same depth of feedback as for HoeGekIsNL but with focus on the elderly and on the issues that are important to them.

• Developed a web application (called Wegweis) to automatically generate personalized advice for schizophrenia patients based on information from their electronic medical records. The application implements an ontology-based approach that links assessment questionnaires with topics from an advice database using concepts from a custom ontology. Robust algorithms exploit the relationships embedded in the ontology to find advice even when there is no direct match. The resulting output is ranked by relevance in a way that was found to be close to how a clinician would prioritize information, given the same data.• Developed an R-package (called Autovar, http://www.autovar.nl/ ) with a web application front-end to automate the process of generating vector autoregression models from electronic patient diary (time series) data. The application also outputs Granger causality summaries, identifying significant causal relationships between selected variables. These results provide insight to clinicians with respect to the efficacy of treatment and medication for individual patients.• Supervised the graduate thesis research of three Bachelor students and four Master students• Reviewed manuscripts for the Artificial Intelligence in Medicine (Elsevier) journal• Reviewed submissions for the following conferences: CNSM 2010, CoopIS 2010, SW 2010, WTA 2010, FedCSIS 2011, IEEE ICWS 2011, IEEE CLOUD 2011, CLOSER 2012, IEEE CLOUD 2012, CoopIS 2012, EITES 2012, IC2IT 2012, ICONS 2012, ICSOC 2012, IEEE ICWS 2012, SoSE 2012, WISE 2012, IEEE CLOUD 2013, ESOCC 2013, FSEN 2013, IEEE ICWS 2013• Participated in the following courses: Maastricht School on Data Mining, Publishing in English, Presenting in English, Software Architecture of Distributed Systems

• Provided in-class support to Bachelor and Master students• Graded weekly lab reports, provided laboratory exams, course syllabi• Assisted the courses Object Oriented Programming, Imperative Programming, Pattern Recognition, Mathematics A, Programming II, Compiler Construction, Computer Architecture & Networks• Worked on improving the automatic code verification system Athena• Wrote texts used on the website of the department of Computer Science for prospective students

• Research in the field of Pattern Recognition and Statistics• Worked on using Shape Vectorization inspired by the V4 Area for Object Recognition