Artificial neural networks in cognitive diagnostic

• Grant number: 18/21063-4
• Support Opportunities: Scholarships abroad - Research
• Start date: October 10, 2018
• End date: July 31, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Mariana Cúri
• Grantee: Mariana Cúri
• Host Investigator: Suely Oliveira
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Institution abroad: University of Iowa, United States
• Associated research grant: 13/07375-0 - CeMEAI - Center for Mathematical Sciences Applied to Industry, AP.CEPID

Abstract

A new model of deep neural network is proposed based on a modification of the variational autoencoder method. The novel propose consists in using the content specification assessed in each item as information for the network conections. This is an important and actual theme in the areas of Educational Data Mining and Psychometrics, in order to enhance cognitive assessment methods in the bigdata era. The impact of this research goes beyond the application of Statistical Learning methods to educational data. One of the aims of this project is to bridge the Computer Science and Statistics notations and concepts related to deep neural network and item response theory models.The project counts on the collaboration of Alina Von Davier, vice-president of ACTNext, professor of Fordham University and ex-director of the Center for Advanced Psychometrics at ETS, and Suely Oliveira, professor at the department of Computer Science at The University of Iowa with expertise in parallel computing and bigdata. The combination of Alina´s experience in the testing industry and research in Psychometrics with the high performance computing expertise of Suely will contribute substantially to my proposed research. The challenge is to succeed in obtaining an efficient ANN method to model performance based on cognitive assessments. It should be emphasized that the research work colaboration with ACT, as an educational industry of testing, goes with the propose of CEPID-CEMEAI (Centro de Ciências Matemáticas Aplicadas a Industria), which I participate as associated researcher.