Barnard, E2012-02-152012-02-152009-11Barnard, E. The challenges of ignorance. 20th Annual Symposium of the Pattern Recognition Association of South Africa (PRASA), Stellenbosch, South Africa, 30 November-01 December 2009, pp 7-10978-0-7992-2356-9http://www.prasa.org/proceedings/2009/prasa09-02.pdfhttp://hdl.handle.net/10204/557120th Annual Symposium of the Pattern Recognition Association of South Africa (PRASA), Stellenbosch, South Africa, 30 November-01 December 2009The authors have previously argued that the infamous "No Free Lunch" theorem for supervised learning is a paradoxical result of a misleading choice of prior probabilities. Here, they provide more analysis of the dangers of uniform densities as ignorance models, and point out the need for a framework that allows for prior probabilities to be constructed in a more principled fashion. Such a framework is proposed for the task of supervised learning, based on the trend of the Bayes error as a function of the number of features employed. Experimental measurements on a number of standard classification tasks confirm the representational utility of the proposed approach.enIgnorance modelsSupervised learningBayes errorConference PresentationBarnard, E. (2009). The challenges of ignorance. PRASA. http://hdl.handle.net/10204/5571Barnard, E. "The challenges of ignorance." (2009): http://hdl.handle.net/10204/5571Barnard E, The challenges of ignorance; PRASA; 2009. http://hdl.handle.net/10204/5571 .TY - Conference Presentation AU - Barnard, E AB - The authors have previously argued that the infamous "No Free Lunch" theorem for supervised learning is a paradoxical result of a misleading choice of prior probabilities. Here, they provide more analysis of the dangers of uniform densities as ignorance models, and point out the need for a framework that allows for prior probabilities to be constructed in a more principled fashion. Such a framework is proposed for the task of supervised learning, based on the trend of the Bayes error as a function of the number of features employed. Experimental measurements on a number of standard classification tasks confirm the representational utility of the proposed approach. DA - 2009-11 DB - ResearchSpace DP - CSIR KW - Ignorance models KW - Supervised learning KW - Bayes error LK - https://researchspace.csir.co.za PY - 2009 SM - 978-0-7992-2356-9 T1 - The challenges of ignorance TI - The challenges of ignorance UR - http://hdl.handle.net/10204/5571 ER -