Abstract
An over-zealous machine learner can automatically generate large, intricate, theories which can be hard to understand. However, such intricate learning is not necessary in domains that lack complex relationships. A much simpler learner can suffice in domains with narrow funnels; i.e. where most domain variables are controlled by a very small subset. Such a learner is TAR2: a weighted-class minimal contrast-set association rule learner that utilizes confidence-based pruning, but not support-based pruning. TAR2 learns treatments; i.e. constraints that can change an agent’s environment. Treatments take two forms. Controller treatments hold the smallest number of conjunctions that most improve the current state of the system. Monitor treatments hold the smallest number of conjunctions that best detect future faulty system behavior. Such treatments tell an agent what to do (apply the controller) and what to watch for (the monitor conditions) within the current environment. Because TAR2 generates very small theories, our experience has been that users prefer its tiny treatments. The success of such a simple learner suggests that many domains lack complex relationships.
Similar content being viewed by others
References
Abts C, Clark B, Devnani-Chulani S, Horowitz E, Madachy R, Reifer D, Selby R, Steece B (1998) COCOMO II model definition manual. Tech. rep., Center for software engineering, USC, http://sunset.usc.edu/COCOMOII/cocomox.html#downloads
Agrawal R, Srikant R (1994) Fast algorithms for mining association rules. In: Proceedings of the 20th international conference on very large databases, available from http://www.almaden.ibm.com/cs/people/ragrawal/papers/vldb94_rj.ps
Bay S, Pazzani M (1999) Detecting change in categorical data: mining contrast sets. In: Proceedings of the fifth international conference on knowledge discovery and data mining, available from http://www.ics.uci.edu/~pazzani/Publications/stucco.pdf
Breiman L, Friedman JH, Olshen RA, Stone CJ (1984) Classification and regression trees. Tech. rep., Wadsworth International, Monterey, CA
Cai C, Fu A, Cheng C, Kwong W (1998) Mining association rules with weighted items. In: Proceedings of international database engineering and applications symposium (IDEAS 98), available from http://www.cse.cuhk.edu.hk/~kdd/assoc_rule/paper.pdf
Clancy D, Kuipers B (1997) Model decomposition and simulation: a component based qualitative simulation algorithm. In: Proceedings of the 14th National Conference on Artificial Intelligence (AAAI-97), AAAI/MIT Press, 1997
Crawford J, Baker A (1994) Experimental results on the application of satisfiability algorithms to scheduling problems. In: AAAI, vol 2. pp 1092–1097
DeKleer J (1986) An assumption-based TMS. Artif Intell 28: 163–196
Feather M, Menzies T (2002) Converging on the optimal attainment of requirements. In: IEEE joint conference on requirements engineering ICRE’02 and RE’02, 9–13th September, University of Essen, Germany, available from http://menzies.us/pdf/02re02.pdf
Feather M, Cornford S, Larson T (2000) Combining the best attributes of qualitative and quantitative risk management tool support. In: 15th IEEE international conference on automated software engineering, Grenoble, France, pp 309–312
Holte R (1993) Very simple classification rules perform well on most commonly used datasets. Mach Learn 11:
Hu Y (2003) Treatment learning: implementation and application. Masters Thesis, Department of Electrical Engineering, University of British Columbia
Josephson J, Chandrasekaran B, Carroll M, Iyer N, Wasacz B, Rizzoni G (1998) Exploration of large design spaces: an architecture and preliminary results. In: AAAI ’98, available from http://www.cis.ohio-state.edu/~jj/Explore.ps
Kohavi R, John GH (1997) Wrappers for feature subset selection. Artif Intell 97(1–2):273–324, URL citeseer.nj.nec.com/kohavi96wrappers.html
Liu B, Hsu W, Ma Y (1998) Integrating classification and association rule mining. In: KDD, pp 80–86, available from http://citeseer.nj.nec.com/liu98integrating.html
Lutz R, Woodhouse R (1999) Bi-directional analysis for certification of safety-critical software. In: 1st International software assurance certification conference (ISACC’99), available from http://www.cs.iastate.edu/~rlutz/publications/isacc99.ps
Madachy R (1997) Heuristic risk assessment using cost factors. IEEE Software 14(3): 51–
Menzies T, Compton P (1997) Applications of abduction: hypothesis testing of neuroendocrinological qualitative compartmental models. Artif Intell Med 10:145–175, available from http://menzies.us/pdf/96aim.pdf
Menzies T, Cukic B (2000a) Adequacy of limited testing for knowledge based systems. International Journal on Artificial Intelligence Tools (IJAIT) Available from http://menzies.us/pdf/00ijait.pdf
Menzies T, Cukic B (2000b) When to test less. IEEE Software 17(5):107–112, available from http://menzies.us/pdf/00iesoft.pdf
Menzies T, Hu Y (2001a) Constraining discussions in requirements engineering. In: First international workshop on model-based requirements engineering, available from http://menzies.us/pdf/01lesstalk.pdf
Menzies T, Hu Y (2001b) Reusing models for requirements engineering. In: First international workshop on model-based requirements engineering, available from http://menzies.us/pdf/01reusere.pdf
Menzies T, Hu Y (2002) Agents in a wild world. In: Rouff C (ed) Formal approaches to agent-based systems, book chapter, available from http://menzies.us/pdf/01agents.pdf
Menzies T, Hu Y (2003) Data mining for very busy people. In: IEEE Computer, available from http://menzies.us/pdf/03tar2.pdf
Menzies T, Singh H (2001) Many maybes mean (mostly) the same thing. In: 2nd International workshop on soft computing applied to software engineering (Netherlands), February, available from http://menzies.us/pdf/00maybe.pdf
Menzies T, Singh H (2003) Many maybes mean (mostly) the same thing. In: Madravio M (ed) Soft computing in software engineering, Springer-Verlag, available from http://menzies.us/pdf/03maybe.pdf
Menzies T, Sinsel E (2000) Practical large scale what-if queries: Case studies with software risk assessment. In: Proceedings ASE 2000, available from http://menzies.us/pdf/00ase.pdf
Menzies T, Easterbrook S, Nuseibeh B, Waugh S (1999) An empirical investigation of multiple viewpoint reasoning in requirements engineering. In: RE ’99, available from http://menzies.us/pdf/99re.pdf
Menzies T, Cukic B, Singh H, Powell J (2000) Testing non determinate systems. In: ISSRE 2000, available from http://menzies.us/pdf/00issre.pdf
Parkes A (1999) Lifted search engines for satisfiability. citeseer.nj.nec.com/parkes99lifted.html
Quinlan R (1992) C4.5: Programs for machine learning. Morgan Kaufman, ISBN: 1558602380
Rymon R (1993) An SE-tree based characterization of the induction problem. In: International conference on machine learning, pp 268–275
Rymon R (1994) An se-tree-based prime implicant generation algorithm. In: Annals of Math and A. I., special issue on Model-Based Diagnosis, vol 11, available from http://citeseer.nj.nec.com/193704.html
Singer J, Gent IP, Smaill A (2000) Backbone fragility and the local search cost peak. J Artif Intell Res 12:235–270, URL citeseer.nj.nec.com/singer00backbone.html
Wang K, He Y, Cheung D, Chin F (2001) Mining confident rules without support requirement. In: 10th ACM International conference on information and knowledge management (CIKM 2001), Atlanta, available from http://www.cs.sfu.ca/~wangk/publications.html
Webb G (2000) Efficient search for association rules. In: Proceeding of KDD-2000 Boston, MA, available from http://citeseer.nj.nec.com/webb00efficient.html
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Menzies, T., Hu, Y. Just enough learning (of association rules): the TAR2 “Treatment” learner. Artif Intell Rev 25, 211–229 (2006). https://doi.org/10.1007/s10462-007-9055-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10462-007-9055-0