Developing Attention Focus Metrics for Autonomous Hypothesis Generation in Data Mining
When facing a data mining task, human experts tend to be responsible for proposing the hypotheses that lead to the discovery of interesting patterns. Recently, there is interest in automating the hypothesis generation process to reduce the load on the human expert during data mining. However, if we want an artificial agent to undertake this new role, we also need new metrics to measure the success of the hypothesis generation mechanism. This paper explores the design of metrics for evaluating hypothesis generation algorithms in terms of differences in the way they focus attention in the data mining search-space. We demonstrate our new metrics applied to three stochastic search based prototype hypothesis generation algorithms. Results show that some differences in attention focus can be identified using our metrics. Directions for further work in attention focus metrics and hypothesis generation algorithms are discussed.
KeywordsHypothesis generation data mining evolutionary computation
Unable to display preview. Download preview PDF.
- 1.Fayyad, U., Piatetsky-Shapiro, G., Smyth, P.: From Data Mining to Knowledge Discovery in Databases. AI Magazine 17, 37–54 (1996)Google Scholar
- 5.Foner, L.N., Maes, P.: Paying Attention to What’s Important: Using Focus Attention to Improve Unsurpervised Learning. In: Proceedings of The Third International Conference on the Simulation of Adaptive Behaviour, pp. 1–20 (1994)Google Scholar
- 7.Graziano, V., Glasmachers, T., et al.: Artificial Curiosity for Autonomous Space Exploration. Acta Futura, 1–16 (2011)Google Scholar
- 8.Goldberg, D.E.: Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley (1989)Google Scholar
- 9.Baluja, S.: Population-Based Incremental Learning: A Method for Integrating Genetic Search based Function Optimization and Competitive Learning. Studies in Fuzziness and Soft Computing 170, 105–129 (1994)Google Scholar
- 10.Price, K.V., Storn, R.M., Lampinen, J.A.: Differential Evolution: A Practical Approach to Global Optimazation. Springer (2005)Google Scholar
- 11.Abbass, H.A.: The Self-Adaptive Pareto Differential Evolution Algorithm. In: Proceedings of the 2002 Congress on Evolutionary Computation, pp. 831–836 (2002)Google Scholar