ALE: Additive Latent Effect Models for Grade Prediction

Abstract

The past decade has seen a growth in the development and deployment of educational technologies for assisting college-going students in choosing majors, selecting courses and acquiring feedback based on past academic performance. Grade prediction methods seek to estimate a grade that a student may achieve in a course that she may take in the future (e.g., next term). Accurate and timely prediction of students' academic grades is important for developing effective degree planners and early warning systems, and ultimately improving educational outcomes. Existing grade prediction methods mostly focus on modeling the knowledge components associated with each course and student, and often overlook other factors such as the difficulty of each knowledge component, course instructors, student interest, capabilities and effort.In this paper, we propose additive latent effect models that incorporate these factors to predict the student next-term grades. Specifically, the proposed models take into account four factors: (i) student's academic level, (ii) course instructors, (iii) student global latent factor, and (iv) latent knowledge factors. We compared the new models with several state-of-the-art methods on students of various characteristics (e.g., whether a student transferred in or not). The experimental results demonstrate that the proposed methods significantly outperform the baselines on grade prediction problem. Moreover, we perform a thorough analysis on the importance of different factors and how these factors can practically assist students in course selection, and finally improve their academic performance.

A Comparative Study of Student Performance Prediction using Pre-Course Data

“…Their results showed that both proposed approaches outperformed existing traditional methods, and course-recommendations based on regression achieved the best results compared to (CSMF) and linear regression. (Ren et al, 2018) proposed student grade prediction model based on the Additive Latent Effect (ALE) within the framework of matrix factorization (MF) that focused on outsourced factors rather than data associated with courses and students. The dataset was obtained from George Mason University and covered the period of Fall 2009 to Spring 2016.…”

“…Machine learning was also utilized to build a recommended system that used matrix factorization and linear regression model for grade prediction (Polyzou and Karypis, 2016). Additive latent affect (ALE) along with matrix factorization (MF) were used to build a student grade prediction model (Ren et al, 2018). In (Acharya and Sinha, 2014), authors have introduced students' predictions of students' performance using machine learning techniques by studying a set of attributes.…”

Predicting Student Performance and Its Influential Factors Using Hybrid Regression and Multi-Label Classification

“…However, grade predictions were less plausible when a student-coursespecific approach was applied. Other works incorporated additive latent effect models (i.e., ALE) within matrix factorization methods to calculate future course grades for the next terms [41]. The proposed ALE models incorporate four factors related to instructors, student academic level, student interest, and knowledge to predict future grades.…”

“…• Single unsupervised learning model: use of only the self-organizing map for multi-label classification. [11], [27], [41], [43], [45], [50], [51]) and their Research Gaps (denoted as RG).…”

Prediction of learning outcomes via clickstream data using machine learning

“…Within a course, achievementoriented learners can often be seen maximizing assessment grades while mastery-oriented learners may undertake more practices and review activities. As learners can make use of the flexibility granted in tertiary education programmes to achieve these goals, models developed for learning analytics can analyze historical records to help learners understand whether they would successfully navigate the programme after deviating from the rec-Chapter 2: Review of Data Mining Approaches for Identifying Learning Behaviors ommended curriculum [26]. While there are many benefits to employ learning analytics to aid learners, it remains debatable if learners should have access to such information as learners may also be discouraged from pursuing their learning if these models constantly provide negative feedback [27,28].…”