Submission to the NIPS 2016 Workshop on Machine Learning for Education Estimating student proficiency : Deep learning is not the panacea

In theoretical cognitive science, there is a tension between highly structured models whose parameters have a direct psychological interpretation and highly complex, general-purpose models whose parameters and representations are difficult to interpret. The former typically provide more insight into cognition but the latter often perform better. This tension has recently surfaced in the realm of educational data mining, where a deep learning approach to estimating student proficiency, termed deep knowledge tracing or DKT [17], has demonstrated a stunning performance advantage over the mainstay of the field, Bayesian knowledge tracing or BKT [3].

[1]  Kenneth R. Koedinger,et al.  Individualized Bayesian Knowledge Tracing Models , 2013, AIED.

[2]  Joseph E. Beck,et al.  Going Deeper with Deep Knowledge Tracing , 2016, EDM.

[3]  Kenneth R. Koedinger,et al.  Performance Factors Analysis - A New Alternative to Knowledge Tracing , 2009, AIED.

[4]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[5]  R. Atkinson,et al.  An Approach to the Psychology of Instruction. , 1972 .

[6]  Zachary A. Pardos,et al.  KT-IDEM: introducing item difficulty to the knowledge tracing model , 2011, UMAP'11.

[7]  Peter Brusilovsky,et al.  Integrating Knowledge Tracing and Item Response Theory: A Tale of Two Frameworks , 2014, UMAP Workshops.

[8]  Chaitanya Ekanadham,et al.  Back to the basics: Bayesian extensions of IRT outperform neural networks for proficiency estimation , 2016, EDM.

[9]  Leonidas J. Guibas,et al.  Deep Knowledge Tracing , 2015, NIPS.

[10]  Michael C. Mozer,et al.  Automatic Discovery of Cognitive Skills to Improve the Prediction of Student Learning , 2014, NIPS.

[11]  Robert V. Lindsey,et al.  Incorporating Latent Factors Into Knowledge Tracing To Predict Individual Differences In Learning , 2013 .

[12]  John R. Anderson,et al.  Knowledge tracing: Modeling the acquisition of procedural knowledge , 2005, User Modeling and User-Adapted Interaction.

[13]  Lancelot F. James,et al.  Generalized weighted Chinese restaurant processes for species sampling mixture models , 2003 .

[14]  Peter Brusilovsky,et al.  General Features in Knowledge Tracing to Model Multiple Subskills, Temporal Item Response Theory, and Expert Knowledge , 2014, EDM.

[15]  José P. González-Brenes.  Modeling Skill Acquisition Over Time with Sequence and Topic Modeling , 2015, AISTATS.

[16]  Deniz Senturk-Doganaksoy,et al.  Explanatory Item Response Models: A Generalized Linear and Nonlinear Approach , 2006, Technometrics.