Data Compression Measures for Meta-Learning Systems
Journal Title: Annals of Computer Science and Information Systems - Year 2018, Vol 15, Issue
Abstract
An important issue in building predictive models is the ability to quickly assess various aspects of the achievable performance of the model to know what outcome we can expect and how to optimally build the model. As instance selection is one of the preprocessing steps that has to be performed anyway, we can use it to obtain the meta-data descriptors for meta-learning systems. When we only need to estimate the classification accuracy of the model, the compression obtained from instance selection is a good approximator. However, when we need to estimate other performance measures such as the precision and sensitivity then the quality of the estimated performance drops. To overcome this issue we propose a new type of compression measure: the balanced compression which shows high correlation with precision and sensitivity of the final classifiers. We also show that the application of the balanced compression as a meta-learning descriptor allows for precise assessment of the model performance, as proved by the presented experimental evaluation.
Authors and Affiliations
Marcin Blachnik, Mirosław Kordos, Sławomir Golak
Keywords
Related Articles
- EP ID EP569787
- DOI 10.15439/2018F87
- Views 21
- Downloads 0
Assistive Smart, Structured 3D Environmental Information for the Visually Impaired and Blind: Leveraging the INSPEX Concept
Inspired by the abilities of contemporary autonomous vehicles to navigate with a high degree of effectiveness, the INSPEX Project 1 seeks to minaturise the sensing and processing technology involved, to produce devices w...
Assertional Reasoning for Concurrent and Communicating BPEL-like Programs
This paper studies verification of programs similar to BPEL4WS (BPEL), the latter being a de facto standard for the web services composition and orchestration. Traditionally, in verification of concurrent and distributed...
"Passeport Vacances": an assignment problem with cost balancing
asseport Vacances is an offer for school-aged children to discover a set of activities during holidays. For more than 30 years, it has been an established social function in several countries, including Germany and Switz...
Reading is Vital, but will it be Invisible? Screens vs. Paper on Our Way to Naturalized Technology of Reading
This analytical position paper aims to open discussion on the future naturalized technology of reading. Our analysis contributes to the discussion that scholars in human-computer interaction should borrow from other disc...
A new task scheduling approach based on Spacing Multi-Objective Genetic algorithm in cloud
The dazzling progress in information and communication technologies, contributed significantly to the emergence of cloud computing paradigm, where it promotes prosperity in all fields of human activity, especially in busin...