Dr. Frederik Schadd

[Mediaan Proejct] Optimization of strategic long-term pilot planning.

Project for Deutsche Telekom: data science platform for the development of virtual digital agents - Development of collaborative data/dialogue labelling tool - Creation of dynamic Rasa-based machine learning NLU model building pipelines - Creation of custom NLU models (tensorflow, scikit-learn, GloVe embeddings)

[Mediaan Projects] Data Science team lead (5 people) Supervision of junior data-scientists for projects involving: - Length-of-stay prediction - Re-admission prediction - Employee intake/outtake forecasting - Patient-journey localization and alerting Other projects/roles: - Introduction of modern standard-practices among data science team (Agile, code review&standard, best practices) - Dutch Privacy Law Compliance: Log analysis, tracking and dashboard visualization of EPD openings

[Mediaan Projects] - Patient intake and bed capacity forecasting for COVID19 patients during outbreak of 2020 pandemic - Patient intake forecasting of Emergency Care Centre - Participation in cross-organizational PoC involving TNO, CBS and CZ for the development and trial of privacy-proof cross-organizational analytics platform - PoC for Sick-leave forecasting for departments/organizational-units

[Mediaan Projects] - Project: Data Science platform for the development of virtual digital agents (AI Center of Excellence) - Project: Predictive Customer Routing and Trigger Campaigns (PK2 Predictive Servicing)

Team member of the analytics team of the Philips CDP2 - Development of churn models based on click-stream data stemming from connected digital propositions. - Churn driver extraction per user to be used as action recommendations. - Big Data analytics and segmentation of clickstream and device data. - Coaching of junior staff members for a project involving churn modeling and advanced analytics. - Development of data pipeline for the purpose of dashboarding

Consultancy work for Philips Research and CDP2. (See below for details)

Team member of the analytics team of the Philips group Connected Digital Plat- forms & Propositions (CDP2) - Realization of ELK+RabbitMQ analytics platform. - Realization of various components of the CDP2 Analytics data science plat- form (Gateway micro-service, API signing, de-duplication, stream ingestion) - Co-ordination of work eorts among several teams spread globally (Eind- hoven, Amsterdam, Bangalore). - Management of AWS-based data visualization platform.

Project member of two internal research projects: - Prototype development of a behavioral change engine, capable of collecting event data of a user and predicting user action based on action history using machine learning techniques. - Development of an insight-rule engine for an infant activity monitor proto- type. Analysis is automated via insight rules, with each rule detecting a specifc pattern in the event data and formulating a user-feedback prompt based on the rule results.

Software development(Matlab, Java) for medical research purposes. Specific Projects: Analysis of spacial electrocardiogram data originating from the atrium. (Matlab) 3-D surface & tissue reconstruction using 2-D MRI data. (Matlab) Linking a Hudson buildserver to an autonomous audio-visual notifier (Nabaztag) by writing a small Mac OS X server. (Java)

Dissertation title: 'Ontology Mapping with Auxiliary Resources'

Graduated with the distinction 'cum laude'. Thesis title: "Monte-Carlo Search Techniques in the Modern Board Game Thurn and Taxis"

Thesis Title: "Hierarchical Opponent Models for Real-Time Strategy Games" Thesis reasearch has been published in Proceedings of the 8th International Conference on Intelligent Games and Simulation (GAMEON’2007) with the title "Opponent modeling in real-time strategy games"