Dr. Adam W. Gripton PhD, MA, BA (Hons.) 4 Stewart Avenue, Currie EH14 5SQ -- Tel (07753) 458495 awgripton@hotmail.com – http://madpaint.org/awg Current Position May 2009 – Present Postdoctoral Research Associate in Image Analysis School of Mathematical and Computer Sciences; Institute of Petroleum Engineering Heriot-Watt University, Edinburgh Summary: My research for this project has involved quantification of statistical model uncertainty, in order to study the relationship between observed noise and the shape of the a posteriori density function of model parameters. We have developed relevant codes for novel Monte Carlo techniques to allow application on a sponsor's external system. Education and Qualifications Doctor of Philosophy (Ph.D.) October 2005 – May 2011 School of Engineering and Physical Sciences Heriot-Watt University, Edinburgh Thesis: Geometric margin domain description with instance-specific margins Summary: My thesis concerns classification of abstract datapoints via the fitting of a spherical domain about the dataset with minimal radius in a way that can deal with missing data without arbitrary imputation and can operate rigorously in a kernel space. Master of Arts, Honour School of Mathematics Jesus College, University of Oxford Grade achieved 2:1 September 2002 – July 2005 Elective units included: Probability, Mathematical Modelling, Applied Analysis, PDEs, Fluid Dynamics, Financial Derivatives, Functional Programming and Data Structures GCE 'A' Levels Thorpe St. Andrew Sixth Form College, Norwich Mathematics A, Physics A, German A September 2000 – June 2002 Research Interests and Experience My research interests are in the fields of data analysis, machine learning and statistical parameter inference. My thesis project related particularly to classification, Bayesian frameworks and kernel machines, with my postdoctoral research concerning uncertainty quantification and Monte Carlo sampling techniques. My PhD project was an EPSRC sponsored studentship supervised by Dr. Weiping Lu. My studies gave me a grounding in classification methods, support vector machines and kernel induced feature spaces, as well as requiring me to use techniques of Bayesian inference to build classification machines to operate on a large, complex dataset provided by an external sponsor. Through my research, I also developed vital skills in programming, computational mathematics and algorithm design, particularly using the Maple and Matlab platforms; I also developed my initiative and creativity skills in finding and exploring novel theory for my thesis topic. Aside from experimental and theoretical work, I was required to design and give presentations of my findings to groups of delegates from the sponsor company and liaise with them in matters concerning project direction. My postdoctoral assignments have required the use of inference techniques, principally based on Markov Chain and Hamiltonian Monte Carlo sampling. My studies have required statistical expertise when dealing with Bayesian inference and autocorrelation of samples; I have also produced codes for Riemann solvers in fluid dynamics for specific application of our methods. More generally, my postdoctoral experience has honed my skills of independent research and has allowed me the scope to guide the direction of project work. My communication skills have been required through our collaboration with two separate research groups, one with a theoretical development remit and the other applying our methods to specific problems. The latter group has also required me to formulate solutions to allow our developed codes to function properly on their systems. In future, as well as developing my skills in the areas with which I am familiar, I hope to direct the path of my research towards further areas of mathematics and computer science that also interest me, including those of game theory, stochastic models, graph theory, logistical systems and operational research. Publications A. Gripton and W. Lu, Kernel Domain Description with Incomplete Data: Using Instance-Specific Margins to Avoid Imputation, International Conference on Pattern Recognition, (2010), pp. 2921-2924 Thesis: A. Gripton, Geometric Margin Domain Description with Instance Specific Margins, PhD thesis, Heriot-Watt University, 2011. Contributed discussion: M. Girolami and B. Calderhead, Riemann Manifold Langevin and Hamiltonian Monte Carlo Methods, Jnl. Royal Stat. Soc. B, 73/2 (2011) pp. 123-214 Paper review: Principal Component Analysis with Interval Imputed missing values, Pattern Recognition Letters (2010) Teaching Experience · · I acted as laboratory co-ordinator for the Praxis Computing Skills unit for firstyear undergraduates for three years (2006-08). In addition to these duties, I engaged in the development of this course alongside the course lecturers, and acted in a supervisory role in my third year for the next year's co-ordinator. I taught a course in Mathematical Physics to third-year undergraduates (2008-9), organising and giving my own classes relating to the material, including sessions of small-group and one-to-one tutoring where necessary, allowing the class to engage with the subject matter and receiving extremely positive feedback. My teaching duties have also included: · · · Calculus B class, 1st year undergraduates, class-based tuition (03/10). Foundation Physics, 1st year undergraduates (2008). Exam invigilation and scribing for students with special needs (Early 07). Conference and Seminar Attendance · · · · · · · · · · · International Conference of Pattern Recognition (ICPR) 2010, Istanbul: Paper proceedings and 3-hour poster discussion session (08/10) Uncertainty in Computer Models (UCM) 2010, Sheffield: Attended inaugural symposium hosted by MUCM research group (07/10) British Applied Mathematics Colloquium (BAMC) 2010, Edinburgh (04/10) Attendance at ERGO lectures in optimisation at University of Edinburgh (11/08) Attendance at Royal Society of Edinburgh lectures (09/08) Heriot-Watt University Research Conference (06/08) MATLAB seminars at Heriot-Watt and Agilent Technologies, Edinburgh (03/08) Presented poster of work completed at Heriot-Watt University Engineering and Physical Sciences Poster Session (18/10/06) Attendance at Bayesian Methods workshop, Heriot-Watt University (07/06) GradSkills Conference "Research Futures", University of St. Andrews (06/06) Attended week-long SUPA Graduate School in Data Analysis, University of Glasgow (01/06) Additional Skills Training Seminars · · · · · · Participated in RDP Creativity and Innovation event (11/10) RDP course in Thesis writing (2009) UK GRAD School Week – PhD training course, Exeter University (07/07) SUPA Entrepreneurship Course (05/07) RDP courses in Plagiarism, Ethics and FOI (11/06) UK GRAD course “How to be an Effective Researcher” (09/06) General Skills · · · · · · Programming: I am skilled in the computational platforms of Maple and MATLAB. Some of my MATLAB codes are available on the public File Exchange. I am familiar with C++, FORTRAN and Haskell. I am currently undertaking a course in JavaScript. Scripting: I have a working knowledge of HTML and CSS, and can design a simple web page to XHTML 1.1 standards. I am currently teaching myself PHP and MySQL with a view to designing an online database application. Computing: I am experienced with MS Windows including console based commands; I have basic Unix/Linux skills and am experienced with IT problem solving. Editing: I am skilled in LaTeX typesetting and familiar with MetaPost vector graphics. I am also experienced with Office applications including Word, Excel and PowerPoint. My typing and proofreading skills are also of a high standard. Languages: English (mother tongue), German (proficient), French (basic), Japanese (basic). Other: Full driving licence since January 2007. References available on request