Learning Curve Extrapolation using Machine Learning

Benefits and Limitations of using LCPFN for Learning Curve Extrapolation

Bachelor Thesis (2024)
Author(s)

P. Johari (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Contributor(s)

Tom Julian Viering – Mentor (TU Delft - Pattern Recognition and Bioinformatics)

O.T. Turan – Mentor (TU Delft - Pattern Recognition and Bioinformatics)

H.S. Hung – Graduation committee member (TU Delft - Pattern Recognition and Bioinformatics)

Faculty
Electrical Engineering, Mathematics and Computer Science
Copyright
© 2024 Pratham Johari
More Info
expand_more
Publication Year
2024
Language
English
Copyright
© 2024 Pratham Johari
Graduation Date
01-02-2024
Awarding Institution
Delft University of Technology
Project
['CSE3000 Research Project']
Programme
['Computer Science and Engineering']
Faculty
Electrical Engineering, Mathematics and Computer Science
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

This study explores the extrapolation of learning curves, a crucial aspect in evaluating learner performance with varying dataset sample sizes. We use the Learning Curve Prior Fitted Network (LC-PFN), a transformer pre-trained on synthetic data with proficiency in approximate Bayesian inference, to investigate its predictive accuracy using the Learning Curve Database (LCDB). The assessment involves MSE as an error metric, with 2 baselines from previous studies where we see it outperform the baseline in some cases and keep on par in others. Additionally, we scrutinize instances where the LC-PFN model may exhibit shortcomings to identify trends in curve extrapolation failures, offering insights for potential modifications to the training dataset. We see a pattern in learners where LC-PFN performs consistently poorly on, whereas no significant pattern can be seen for datasets.

Files

Pratham_v_1_0_5.pdf
(pdf | 4.35 Mb)
License info not available