Interesting papers related to my recent work and both from http://www.princeton.edu/~harman/Papers/ Title: Statistical Learning Theory as a Framework for the Philosophy of Induction Link: http://www.princeton.edu/~harman/Papers/Induction-JH.pdf Abstract: Statistical Learning Theory (e.g., Hastie et al., 2001; Vapnik, 1998, 2000, 2006) is the basic theory behind contemporary machine learning and data-mining. We suggest that the theory provides an excellent framework for philosophical thinking about inductive inference.1 We begin by sketching certain aspects of statistical learning theory, then comment briefly on philosophical implications for reliability theories of justification, Karl Popper’s (1979, 2002) appeal to falsifiability, simplicity as relevant to inductive inference, and whether direct induction is an instance of inference to the best explanation. Title: Reliable Reasoning: Induction and Statistical Learning Theory Link: http://www.princeton.edu/~kulkarni/Papers/b2006_hk.pdf Abstract: The book arises from an introductory course on Learning Theory and Epistemology we have been teaching jointly in Electrical Engineering and Philosophy at Princeton University. This low-level undergraduate course serves as an introduction to aspects of philosophy, computer science, engineering, statistics and cognitive science. It is open to all students and has no specific prerequisites other than some analytical skills and intellectual curiosity. Although much of the material is technical, we have found that the main points are both accessible to and appreciated by a broad range of students. In each class, our students have included freshmen through seniors, with majors from the sciences, engineering, humanities, and social sciences. We acknowledge with thanks a Curriculum Development Grant for this course from the 250th Anniversary Fund for Innovation in Undergraduate Education from Princeton University and are grateful to the many students who have discussed the content of the course with us. --

※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 140.112.91.4