Research Scientist (Machine Learning) at ScanSafe and Student at University College London, U. of London
London, United Kingdom
- Current
-
- Research Scientist (Machine Learning) at ScanSafe
- Past
-
- Data Matching Analyst / Developer at National Cancer Registry, Ireland
- Software Developer at Vistech
- ITS Staff (Summers 1999, 2000) at Surface Mount Technology Center
- Education
-
- University College London, U. of London
- University College Cork
- Connections
-
12
connections
- Industry
- Computer & Network Security
- Websites
Bryan Feeney’s Summary
Initially a software developer, work at the National Cancer Registry engendered an interest in developing novel methods to deal with large datasets, particularly using machine-learning techniques (having briefly encountered them in my degree). I acted on this interest by pursuing a Masters in Intelligent Systems at University College London, and am now putting such techniques into practice at ScanSafe. Going forward I look to expand and deepen my knowledge of such techniques as I employ them, and ultimately develop novel methods myself.
Bryan Feeney’s Specialties:
Machine Learning, Computer Security, Software Development, Java, C++, C, Delphi, J2EE, EJB3, JSF, Struts, Hibernate 3
Bryan Feeney’s Experience
-
Research Scientist (Machine Learning)
ScanSafe
(Privately Held; 51-200 employees; Computer & Network Security industry)
May 2007 — Present (1 year 3 months)
Using machine learning techniques to develop novel means and methods for early identification of malware threats and further to classify websites as dangerous, inappropriate or benign. Software development involves Java, Hibernate, Spring, Struts and Oracle.
-
Data Matching Analyst / Developer
National Cancer Registry, Ireland
(Government Agency; 51-200 employees; Research industry)
May 2003 — June 2006 (3 years 2 months)
Initially employed as a developer,
Designed and created dynamic website using PHP, XHTML, CSS, SVG and MySQL.
Later developed software and systems to encrypt and securely transmit sensitive data. Developed software included applications and libraries, developed with Delphi using of Win32 for lower-level work such as multithreading within libraries and creating hooks for random number generation
Subsequently worked in software research, investigating means and methods to probabilistically identify one or more records in one or more datasets which referred to the same entity (e.g. a person) despite data-entry errors and omissions. Used in dataset de-duplication and merging. -
Software Developer
Vistech
(Public Company; 11-50 employees; Computer Software industry)
April 2001 — September 2001 (6 months)
Initially worked in quality-assurance, creating and executing test-plans based on functional specfications. Later worked in software development, creating dynamic websites using Java, JSP and MySQL.
-
ITS Staff (Summers 1999, 2000)
Surface Mount Technology Center
(Privately Held; 501-1000 employees; Electrical/Electronic Manufacturing industry)
June 1999 — September 2000 (1 year 4 months)
Jun-Sep 1999: Performed tests to measure Y2K compatibility: carried out software and hardware updates as appropriate, and created reports on complicance. Also provided general user-support
Jun-Sep 2000: Performed additional support and maintenance tasks, and continued to perform user-support.
Bryan Feeney’s Education
-
University College London, U. of London
M.Sc., Intelligent Systems, 2006 — 2007
Graduated with Distinction. Course content included supervised and unsupervised learning, probabilistic inference, information theory, reinforcement learning, and evolutionary learning techniques. Masters thesis involved the development of a kernel-based method to solve the bandit problem with shifting reward distribution (e.g. an adserver identifying ads to display to customers based on previous customer feedback).
-
University College Cork
B.Sc., Computer Science, 1998 — 2002
Graduated with first class honours (1.1). Course content included algorithms and automata, computer security, concurrency, databases and artificial intelligence. Final year project investigated the use of genetic algorithms to solve the NP-hard Golomb Ruler problem.
- Activities and Societies:
- Member of the UCC Mountaineering Club. Treasurer in 2002. Acted as leader on walks in Ireland, Scotland and Wales. Coordinated trips to Wales and the West of Ireland.
Additional Information
Bryan Feeney’s Websites:
Bryan Feeney’s Interests:
Mountaineering, Machine Learning, Applications of Machine Learning, Politics
Bryan Feeney’s Groups:
IEEE
Bryan Feeney’s Honors:
Chevy-Chase Bank Scholarship in 1999 on the basis of Leaving Certificate (pre-college) examination results.
During undergraduate degree, was twice declared a "College Scholar" on the basis of end-of-year results.
