.png){kind=icon}
返回
Grey wolf optimization-extreme learning machine for automatic spoken language identification
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-023-14473-3.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Natural language classification and determination based on a particular content and dataset is carried out using Spoken Language Identification (LID) which typically involves the extraction of valuable elements in a mature data processing procedure whereby the regular LID features had been developed using the Mel Frequency Cepstral Coefficient (MFCC), Shifted Delta Coefficient (SDC), Gaussian Mixture Model (GMM) and an i-vector framework. However, there remains a need for optimization in terms of the learning process so as to allow for all the knowledge embedded in the extracted features to be captured completely. A powerful machine learning algorithm known as Extreme Learning Machine (ELM) is used for conducting regression and classification and can train single hidden layer neural networks effectively. Yet, ELM's learning process remains under-optimized owing to the entrenched random weights selection in the input hidden layer. Based on the standard feature extraction, this current study chooses ELM as the learning model for LID. An optimized method known as the Enhanced Self-Adjusting-ELM (ESA-ELM) has been chosen as a benchmark with enhancements via the adoption of an alternate optimization approach i.e., Grey Wolf Optimisation (GWO) rather than Enhanced Ameliorated Teaching Learning-Based Optimization (EATLBO) to ensure higher performance. Ultimately, this enhanced version of the ESA-ELM is referred to as a Grey Wolf Optimisation-Extreme Learning Machine (GWO-ELM). The results generation is carried out based on LID using the exact benchmark dataset that was derived from eight separate languages. The results indicated that the GWO-ELM LID has a much superior performance than the ESA-ELM LID with respective accuracies of 100.00% for the former and merely 96.25% for the latter.
Keyword:
Extreme learning machine
Grey wolf optimisation
Language identification
期刊
IF:
3
论文数:
2.0W
被引数:
3.2W
机构
引用论文
Productivity enhancement of solar still by PCM and Nanoparticles miscellaneous basin absorbing materials.png)
Desalination
OAAI
DesalinationIF0
OAAI分享收藏分享收藏Quantitative 201thallium scintigraphy for prediction of histological response to neoadjuvant chemotherapy in osteosarcoma; systematic review and meta-analysis定量201铊闪烁显像预测骨肉瘤新辅助化疗的组织学反应；系统评价和Meta分析
PREAI
分享收藏分享收藏
查看更多内容