ProbabilisticClassificationLearning

An ability which seems to be dependent on the basal ganglia Knowlton et al '96. Given some cues, you’re supposed to predict a target attribute. You get a a bunch of training instances (cues paired with the target), and then are tested. The canonical task is the WeatherPredictionTask.

A double dissociation was shown between hippocampal and basal ganglia learning systems using this task; hippocampal amnesiacs had trouble remembering the training episodes, but learned the classification O.K., while Parkinson's patients remembered the training episodes, but had deficiencies in learning the classification Knowlton et al '96.

In addition, patients who had had a [[pallidotomy?]] had more deficiencies; pallidotomy (combined with Parkinson’s) appeared to hinder the most weakly associated cues Sage et al '03.

This sort of task is what would be called, in A.I., a AIWiki:SupervisedLearningTask.

Refs

• Barbara J. Knowlton, Jennifer A. Mangels, and Larry R. Squire. Neostriatal Habit Learning System in Humans. Science 1996; 273 (5280) : 1399-1402. (in Reports)
• Sage, Jennifer R., Anagnostaras, Stephan G., Mitchell, Shawn, Bronstein, Jeff M., De Salles, Antonio, Masterman, Donna, Knowlton, Barbara J. Analysis of Probabilistic Classification Learning in Patients With Parkinson's Disease Before and After Pallidotomy Surgery. Learn. Mem. 2003 10: 226-236 (2003)

CategoryLearning

CategoryBasalGanglia