Journal article
A modified Lanczos Algorithm for fast regularization of extreme learning machines
Neurocomputing (Amsterdam), Vol.414, pp.172-181
11/13/2020
DOI: 10.1016/j.neucom.2020.07.015
Abstract
This paper presents a new regularization for Extreme Learning Machines (ELMs). ELMs are Randomized Neural Networks (RNNs) that are known for their fast training speed and good accuracy. Nevertheless the complexity of ELMs has to be selected, and regularization has to be performed in order to avoid underfitting or overfitting. Therefore, a novel Regularization is proposed using a modified Lanczos Algorithm: Iterative Lanczos Extreme Learning Machine (Lan-ELM). As summarized in the experimental Section, the computational time is on average divided by 4 and the Normalized MSE is on average reduced by 11%. In addition, the proposed method can be intuitively parallelized, which makes it a very valuable tool to analyze huge data sets in real-time.
Details
- Title: Subtitle
- A modified Lanczos Algorithm for fast regularization of extreme learning machines
- Creators
- Renjie Hu - ILT Department, University of Houston, Houston, USAEdward Ratner - ILT Department, University of Houston, Houston, USADavid Stewart - Department of Mathematics, The University of Iowa, Iowa City, USAKaj-Mikael Björk - Arcada University of Applied Sciences, Helsinki, FinlandAmaury Lendasse - ILT Department, University of Houston, Houston, USA
- Resource Type
- Journal article
- Publication Details
- Neurocomputing (Amsterdam), Vol.414, pp.172-181
- DOI
- 10.1016/j.neucom.2020.07.015
- ISSN
- 0925-2312
- eISSN
- 1872-8286
- Publisher
- Elsevier B.V
- Language
- English
- Date published
- 11/13/2020
- Academic Unit
- Industrial and Systems Engineering; Mathematics
- Record Identifier
- 9984066342402771
Metrics
37 Record Views