Learning Time Acceleration in Support Vector Regression: A Case Study in Educational Data Mining

Abstract

The development of a country involves directly investing in the education of its citizens. Learning analytics/educational data mining (LA/EDM) allows access to big observational structured/unstructured data captured from educational settings and relies mostly on machine learning algorithms to extract useful information. Support vector regression (SVR) is a supervised statistical learning approach that allows modelling and predicts the performance tendency of students to direct strategic plans for the development of high-quality education. In Brazil, performance can be evaluated at the national level using the average grades of a student on their National High School Exams (ENEMs) based on their socioeconomic information and school records. In this paper, we focus on increasing the computational efficiency of SVR applied to ENEM for online requisitions. The results are based on an analysis of a massive data set composed of more than five million observations, and they also indicate computational learning time savings of more than 90%, as well as providing a prediction of performance that is compatible with traditional modeling.

Multidimensional State Data Reduction and Evaluation of College Students’ Mental Health Based on SVM

“…Nevertheless, the basic mathematical models are still very useful and have eventually been adapted to many computational scenarios. Current research tools are divided into nonparametric and parametric approaches: the nonparametric approach does not require the assumption of knowledge of any parameters and uses nonparametric techniques to estimate the density of the distribution, for example, histograms and Parzen window estimation; the parametric approach requires the assumption that normal data are generated based on parametric distributions, and it requires these parameters from training samples, for example, outlier detection methods based on normal distributions [13]. Neural networks, which may be classified into single-classification neural networks and multiclassification neural networks [14], are an important field of nonlinear modeling approaches.…”

Intelligent Analysis of Exercise Health Big Data Based on Deep Convolutional Neural Network

“…The study of student physical health big data started earlier in many developed countries such as the United States. Although the research is in the initial stage of development, with the introduction of a series of national guidelines and policies, the research results are becoming increasingly fruitful [ 6 ]. DeepPose defines the human posture estimation as a nodal point regression problem, using convolutional neural networks to directly regress the coordinates of each node, and the authors use the idea of the multistage cascade.…”