Pedram, Sami, Suyog, Kashyap, Aurelién, Fabio, Petrus, Mert.
Discussed the teaching problem and how to design the best item to recommend to the user. The utility should be a compromise between accuracy and forgetting.
|Mert||Gave a short long update. Talked about iterative machine teaching and its relation to modelling made easy.|
Additional Material by Suyon
Using Bayes Nets to help users converge on the right representation
Bramley, N. R., Dayan, P., Griffiths, T. L., & Lagnado, D. A. (2017). Formalizing Neurath’s ship: Approximate algorithms for online causal learning. Psychological review, 124(3), 301
Extends the rational model of categorization to (iterated) function learning
Lucas, C. G., Griffiths, T. L., Williams, J. J., & Kalish, M. L. (2015).A rational model of function learning. Psychonomic bulletin & review, 22(5), 1193-1215.
GP based machine teaching / optimal experiment design where they avoid being constrained by parametric assumptions
Chang, J., Kim, J., Zhang, B. T., Pitt, M. A., & Myung, J. I. Modeling Delay Discounting using Gaussian Process with Active Learning.
Some preliminary work in embedding the GCM exemplar model of categorization within a machine learning context to applied settings:
i) Using deep-learning to estimate the similarity matrices used by the GCM model
Sanders, C., & Nosofsky, R. M. (2018). Using Deep-Learning Representations of Complex Natural Stimuli as Input to Psychological Models of Classification. In CogSci.
ii) Using the GCM model in an optimal teaching scenario
Nosofsky, R. M., Sanders, C. A., Zhu, X., & McDaniel, M. A. (2019). Model-guided search for optimal natural-science-category training exemplars: A work in progress. Psychonomic bulletin & review, 26(1), 48-76.
- Next week: Petrus and Mert provide longer updates next week.