Team‎ > ‎

Matthew Walker

Position: Data Scientist and Developer

Email address: matthew@bodekerscientific.com

Academic Qualifications:

·        1999 BInfSc, Massey University, Computer Science and Mathematics

·        2000 BInfSc (Hons, 1st class), Massey University, Computer Science

·        2007 PhD, Massey University, Computer Science (Developing and Evaluating Incremental Evolution using High Quality Performance Measures for Genetic Programming. Supervised by Dr. C. H. Messom)

Positions:

·        Self-Employed Consulting.  2018–2021.  Developed audio conferencing applications for synchronised singing; analysis for a New Zealand Hospital that uncovered the benefits of treatment-induced erythema (burning) in both treatment efficacy and patient withdrawal rates; provided second opinion and re-analysis of internal consistency for an online metal health questionnaire; created simulation of team dynamics including visualisation and its results

·        Manager. 2016–2018.  Peter Walker Consultants, New Zealand.  Wrapped up two in-progress contracts: the completion of an automated beverage-production machine, and the design, manufacture, and commissioning of a plant to control the pH of a waste stream. As well as the technical requirements associated with taking over the contracts, the role included significant management: of the client relationships, of the subcontractors, and of the scheduling of the work around the closure of the company.

·        Data Scientist / Software Developer. 2016.  Motivated to reduce the number of deaths caused by tourists in head-on collisions, we started investigating the feasibility of producing a vision-based in-car detector to warn drivers when they crossed to the wrong side of the road. We successfully created a proof-of-concept using TensorFlow on a RaspberryPi, with training done on AWS EC2 servers.

·        Artificial Intelligence and Machine Learning Course Designer.  2016.  Tech Futures Lab, New Zealand. Created course content for a 10-week long, 40 hours/week, deep-dive programme in Artificial Intelligence and Machine Learning. Collaborated significantly with the Robotics programme to form novel, practical joint projects using Python.

·        Software Developer / Data Scientist. 2014–2016.  Direct Control, New Zealand.  Responsible for automated analytics implementation for a client with a portfolio of many high-rise buildings. Developed models of energy profiles, for buildings and plant, based on analysis of historic data. Implemented automated fault-detection for energy savings.  Part of the development and commissioning team for a building management system (BMS) contract for a large data-centre. Developed and documented modules to analyse power distribution and redundancy and to calculate power usage efficiency. Introduced automated software testing. Introduced version control using Git and developed processes for software releases.

·        Junior Consultant 2011–2014.  Peter Walker Consultants, New Zealand.  Implemented Monte-Carlo financial model for risk assessment for a new Spray-Drying business.  Involved in consulting for the Commerce Commission on the validity of Fonterra's Milk Price Model. Helped reverse-engineer the functionality of Fonterra's highly-complex multi-spreadsheet model; assisted with verification via implementation of an independent model. Responsible for reporting on the original model's quality from a software design perspective.  Part of commissioning team for membrane-filtration plants, one of which, including initial operation and operator training, was in France, interacting with client in French. Programmed web-based interfaces for filtration plants. Helped design and implement a plant intended to run entirely automated 24/7-filtration experiments.

·        Post-Doctorate Researcher 2008–2011.  Université Laval and Thales, Canada.  Worked (in French and English) on the automation of the placement of sensors in Wireless Sensor Networks (WSN) for DRDC (Defence Research and Development Canada) utilising a multi-objective developmental genetic programming approach.

Scholarships:

·        EvoStar Travel Bursary (approx $1000) 2007

·        Foundation for Research, Science and Technology (FRST) Bright Future's Top Doctoral Scholarship (top 5% of PhD candidates; $25,000 plus expenses/year) 2001-2004

·        Institute of Information and Mathematical Sciences Scholarship (approx $4,000) 2000

Awards:

·        Nominated for best paper at EvoStar 2007 for Confidence Intervals for Computational Effort Comparisons. 2007

·        Top Student in Computer Science (NZ Computer Society) 2000

Publications:

·        M. Walker, H. Edwards, and C. Messom. Success Effort and Other Statistics for Performance. Comparisons in Genetic Programming. In Congress on Evolutionary Computing (CEC). 2007.

·        M. Walker, H. Edwards, and C. Messom. The reliability of confidence intervals for computational effort comparisons. In Proceedings of the 9th annual conference on Genetic and evolutionary computation (GECCO), pages 1716-1723, volume 2, 2007.

·        M. Walker, H. Edwards and C. Messom. “Success effort” for performance comparisons. In GECCO '07: Proceedings of the 9th annual conference on genetic and evolutionary computation, volume 2, 2007.

·        M. Walker, H. Edwards and C. Messom. Confidence Intervals for Computational Effort Comparisons. In Genetic Programming: Proceedings of the 10th European Conference, EuroGP 2007. Lecture Notes in Computer Science, volume 4445. 2007.

·        M. Walker. Comparing the performance of incremental evolution to direct evolution. In Second International Conference on Autonomous Robots and Agents (ICARA), pages 119-124, December 2004.

·        M. Walker. Mutated Seeds: A Performance Enhancer for Incremental Evolution. In Proceedings of the Institute of Information and Mathematical Sciences (IIMS) Postgraduate Conference, pages 95-99. 2004.

Comments