analiza aiqbench
This dissertation introduces Seer, a statistical framework for modeling machine learning performance curves that predicts training requirements and asymptotic accuracy limits. By reducing complex nonlinear programming tasks to efficient regression algorithms, the work provides practical tools for evaluating classification-learning constraints across diverse domains like medical diagnosis and agriculture. While the research is decades old, its methodological approach to quantifying learning speed remains relevant for understanding fundamental ML scaling laws.
Seer: Maximum Likelihood Regression for Learning-Speed Curves
arXiv:2610.02610v1 Announce Type: new Abstract: The research presented here focuses on modeling machine-learning performance. The thesis introduces Seer, a system that generates empirical observations of classification-learning performance and then uses those observations to create statistical models. The models can be used to predict the number of training examples needed to achieve a desired level and the maximum accuracy possible given an unlimited number of training examples. Seer advances the state of the art with 1) models that embody the best constraints for classification learning and most useful parameters, 2) algorithms that efficiently find maximum-likelihood models, and 3) a demonstration on real-world data from three domains of a practicable application of such modeling. The first part of the thesis gives an overview of the requirements for a good maximum-likelihood model of classification-learning performance. Next, reasonable design choices for such models are explored. Selection among such models is a task of nonlinear programming, but by exploiting appropriate problem constraints, the task is reduced to a nonlinear regression task that can be solved with an efficient iterative algorithm. The latter part of the thesis describes almost 100 experiments in the domains of soybean disease, heart disease, and audiological problems. The tests show that Seer is excellent at characterizing learning-performance and that it seems to be as good as possible at predicting learning performance. Finally, recommendations for choosing a regression model for a particular situation are made and directions for further research are identified.
Abstract:The research presented here focuses on modeling machine-learning performance. The thesis introduces Seer, a system that generates empirical observations of classification-learning performance and then uses those observations to create statistical models. The models can be used to predict the number of training examples needed to achieve a desired level and the maximum accuracy possible given an unlimited number of training examples. Seer advances the state of the art with 1) models that embody the best constraints for classification learning and most useful parameters, 2) algorithms that efficiently find maximum-likelihood models, and 3) a demonstration on real-world data from three domains of a practicable application of such modeling. The first part of the thesis gives an overview of the requirements for a good maximum-likelihood model of classification-learning performance. Next, reasonable design choices for such models are explored. Selection among such models is a task of nonlinear programming, but by exploiting appropriate problem constraints, the task is reduced to a nonlinear regression task that can be solved with an efficient iterative algorithm. The latter part of the thesis describes almost 100 experiments in the domains of soybean disease, heart disease, and audiological problems. The tests show that Seer is excellent at characterizing learning-performance and that it seems to be as good as possible at predicting learning performance. Finally, recommendations for choosing a regression model for a particular situation are made and directions for further research are identified.
104 pages. Ph.D. dissertation, Department of Computer Science, University of Illinois at Urbana-Champaign, 1995. Original dissertation deposited in arXiv in 2026
Machine Learning (cs.LG)
arXiv:2610.02610 [cs.LG]
(or arXiv:2610.02610v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.02610 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 350 cuvinte
Subiecte conexe
Exploding variance of means of exponentials: least-squares to the rescue – Machine Learning Research Blog
francisbach.com · 27 sept. 2026
Fine-tune a search agent with multi-turn RL on Amazon SageMaker AI
AWS · 2 oct. 2026
Humans Are Not Chinchillas: Revisiting “How Quick and Big Would A Software Intelligence Explosion Be?”
LessWrong · 5 oct. 2026
Would We See It Coming? Preference Falsification Cascades in Multi-Agent Systems
LessWrong · 4 aug. 2026