
Variational Probabilistic Inference and the QMRDT Network
We describe a variational approximation method for efficient inference i...
read it

Robust, Accurate Stochastic Optimization for Variational Inference
We consider the problem of fitting variational posterior approximations ...
read it

Stochastic Variational Inference for Hidden Markov Models
Variational inference algorithms have proven successful for Bayesian ana...
read it

Cauchy noise loss for stochastic optimization of random matrix models via free deterministic equivalents
Based on free probability theory and stochastic optimization, we introdu...
read it

Variational Bayesian Inference for the PolytomousAttribute Saturated Diagnostic Classification Model with Parallel Computing
As a statistical tool to assist formative assessments in educational set...
read it

Asynchronous Stochastic Variational Inference
Stochastic variational inference (SVI) employs stochastic optimization t...
read it

Simultaneous inference of periods and periodluminosity relations for Mira variable stars
The Period–Luminosity relation (PLR) of Mira variable stars is an import...
read it
Scalable Estimation Algorithm for the DINA Qmatrix Combining Stochastic Optimization and Variational Inference
Diagnostic classification models (DCMs) enable finergrained inspection of the latent states of respondents' strengths and weaknesses. However, the accuracy of diagnosis deteriorates when misspecification occurs in the predefined itemattribute relationship, which is defined by a Qmatrix. To forestall misdiagnosis, several Qmatrix estimation methods have been developed in recent years; however, their scalability to largescale assessment is extremely limited. In this study, we focus on the deterministic inputs, noisy "and" gate (DINA) model and propose a new framework for Qmatrix estimation in which the goal is to find the Qmatrix with the maximized marginal likelihood. Based on this framework, we developed a scalable estimation algorithm for the DINA Qmatrix by constructing an iteration algorithm utilizing stochastic optimization and variational inference. The simulation and empirical studies reveal that the proposed method achieves highspeed computation and good accuracy. Our method can be a useful tool for estimating a Qmatrix in largescale settings.
READ FULL TEXT
Comments
There are no comments yet.