A Non-Deterministic Strategy for Searching Optimal Number of Trees Hyperparameter in Random Forest
Journal Title: Annals of Computer Science and Information Systems - Year 2018, Vol 15, Issue
Abstract
In this paper, we present a non-deterministic strategy for searching for optimal number of trees hyperparameter in Random Forest (RF). Hyperparameter tuning in Machine Learning (ML) algorithms is essential. It optimizes predictability of an ML algorithm and/or improves computer resources utilization. However, hyperparameter tuning is a complex optimization task and time wasting. We set up experiments with the goal of maximizing predictability, minimizing number of trees and minimizing time of execution. Compared to the deterministic algorithm, this research's non-deterministic algorithm recorded an average percentage accuracy of approximately 98\\%, number of trees percentage average improvement of 44.64\\%, average time of execution mean improvement ratio of 212.79 and an average improvement of 93\\% iterations. Moreover, evaluations using Jackkife Estimation show stable and reliable results from several experiment runs of the non-deterministic strategy. The non-deterministic approach in selecting hyperparameter shows a significant accuracy and better computer resources (i.e cpu and memory time) utilization. This approach can be adopted widely in hyperparameter tuning, and in conserving utilization of computer resources i.e green computing.
Authors and Affiliations
Kennedy Senagi, Nicolas Jouandeau
Keywords
Related Articles
- EP ID EP569796
- DOI 10.15439/2018F202
- Views 18
- Downloads 0
Clustered Comparative Analysis of Security Sensor Discrimination Data
Security alarm is used to protect from burglary (theft), property damage and from intruders. These security alarms consists sensors and alerting device to indicate the intrusion. Clustering is data mining technique which...
News articles similarity for automatic media bias detection in Polish news portals
Digital media have enormous impact on the public opinion. In the ideal world the news in public media should be presented in a fair and impartial way. In practice the information presented in digital media is often biase...
Automated generator for complex and realistic test data—a case study
Some type of tests, especially stress tests and functional tests, require a large amount of realistic test data. In this paper, we propose a tool JOP (Java Object Populator) that uses a pseudorandom number generator in o...
Analysis of inter-channel dependencies in audio lossless block coding
In this paper the basics of data predictive modeling (using the method of minimization mean square error) for lossless audio compression are presented. The described research focuses on inter-channel analysis and setting...
Static typing and dependency management for SOA
Several problems related to work reliability appear while building service-oriented systems. The first problem consists in lack of static typing and lack of inter-service data type checking. The second one consists in hi...