Bryant, C. H., Muggleton, S. H., Page, C. D. & Sternberg, M. J. E. (1999). Combining active learning with inductive logic programming to close the loop in machine learning.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....express the easy aspects of the task at hand and then collect a small number of training examples to refine and extend this prior knowledge. Finally, we plan to use active learning to allow our ILP systems to select more effective training examples for interactively learning relational concepts [26]. By intelligently choosing the examples for users to label, better extraction accuracy can be obtained from fewer examples, thereby greatly reducing the burden on the users of our ILP systems. 5 Related Work Although it is the most widely studied, ILP is not the only approach to relational data ....

S. Muggleton, C. Bryant, C. Page, and M. Sternberg. Combining active learning with inductive logic programming to close the loop in machine learning. In S. Colton, editor, Proceedings of the AISB'99 Symposium on AI and Scientific Creativity (informal proceedings), 1999.

.... compounds [37] and to model structure activity relations [22] More recently, Muggleton has developed the general purpose ILP language PROGOL, 31] An exciting ongoing project with PROGOL was discussed at the symposium by Chris Bryant from the Department of Computer Science, University of York, [6]. The team have been working 3 towards closed loop scientific discovery, in which experiments are planned automatically and carried out by robots, with the results analysed using ILP techniques. The robotics technology required to achieve this has recently become available, and this is one of the ....

....in mathematics, 9] 15] 3. Spotting examples of a phenomenon, as happened recently with the successful automated identification of distant quasars in astronomy, 21] 4. Designing experiments to test hypotheses and performing closed loop discovery to illustrate the progress of a theory, [6] , 41] 5. Making explicit unquestioned assumptions in a domain, such as those identified about leukemia, 12] There is much real potential for automated discovery programs to produce findings which have a great impact on science. More and more programs are being written to act in creative ....

C. H. Bryant, S. H. Muggleton, C. D. Page, and M. J. E. Sternberg. Combining active learning with inductive logic programming to close the loop in machine learning. In AISB'99 Symposium on AI and Scientific Creativity, pages 59--64, Edinburgh, Scotland, April 1999.

....for use in pharmacophore discovery. C. 6 International Collaboration The proposer has been collaborating with Professor Stephen Muggleton of the University of York on the topic of closed loop learning, in which a machine learning systems proposes experiments that can be carried out by robot [35]. Information from these automated experiments then goes back to the machine learning system, which may propose further experiments before arriving at a nal hypothesis. Professor Muggleton recently received a grant from the British Engineering and Physical Sciences Research Council to study ....

S.H. Muggleton, C.H. Bryant, C.D. Page, and M.J.E. Sternberg. Combining active learning with inductive logic programming to close the loop in machine learning. In S. Colton, editor, Proceedings of the AISB'99 Symposium on AI and Scientic Creativity (informal proceedings), 1999.

No context found.

S. Muggleton, C. Bryant, C. Page, and M. Sternberg. Combining active learning with inductive logic programming to close the loop in machine learning. In S. Colton, editor, Proceedings of the AISB'99 Symposium on AI and Scientific Creativity (informal proceedings), 1999.

No context found.

Muggleton, S.; Bryant, C.; Page, C.; and Sternberg, M. 1999. Combining active learning with inductive logic programming to close the loop in machine learning. In Colton, S., ed., Proceedings of the AISB'99 Symposium on AI and Scientific Creativity (informal proceedings).

No context found.

S. Muggleton, C. Bryant, C. Page, and M. Sternberg. Combining active learning with inductive logic programming to close the loop in machine learning. In S. Colton, editor, Proceedings of the AISB'99 Symposium on AI and Scientific Creativity (informal proceedings), 1999.

....express the easy aspects of the task at hand and then collect a small number of training examples to refine and extend this prior knowledge. Finally, we plan to use active learning to allow our ILP systems to select more effective training examples for interactively learning relational concepts [Muggleton et al..1999]. By intelligently choosing the examples for users to label, better extraction accuracy can be obtained from fewer examples, thereby greatly reducing the burden on the users of our ILP systems. 11 5 Related Work Although it is the most widely studied, ILP is not the only approach to relational ....

Muggleton, S.; Bryant, C.; Page, C.; and Sternberg, M. 1999. Combining active learning with inductive logic programming to close the loop in machine learning. In Colton, S., ed., Proceedings of the AISB'99 Symposium on AI and Scientific Creativity (informal proceedings).

No context found.

Bryant, C. H., Muggleton, S. H., Page, C. D. & Sternberg, M. J. E. (1999). Combining active learning with inductive logic programming to close the loop in machine learning.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC