Joachim Diederich, "Rule Extraction from Support Vector Machines"
2008 | pages: 264 | ISBN: 3540753893 | PDF | 10,3 mb
2008 | pages: 264 | ISBN: 3540753893 | PDF | 10,3 mb
Over a period spanning more than a decade,supportvector machines (SVMs) have evolved into a leading machine learning technique. SVMs are being applied to a wide range of problems, including bioinformatics, face rec- nition, text classi?cation and many more. It is fair to say that SVMs are one of the most important methods used for data mining with a wide range of software available to support their application. A signi?cant barrier to the widespread application of support vector machines is the absence of a capability to explain, in a human comprehensible form,either the process by which anSVM arrivesat a speci?c decision/result, or more general, the totality of knowledge embedded in these systems. This lack of a capacity to provide an explanation is an obstacle to a more g- eral acceptance of “back box” machine learning systems. In safety-critical or medical applications, an explanation capability is an absolute requirement. This book provides an introduction and overviewof methods used for rule extractionfrom support vector machines.The ?rst parto?ers an introduction to the topic as well as a summary of current research issues.
My Links