Scholarlink

<p>Dementia is an irreversible neurological condition that primarily impacts older individuals. The development of effective automated approaches is necessary to achieve timely and accurate detection. Various innovative methodologies have been suggested by researchers to categorize dementia. However, it is crucial to have a thorough comprehension of the current research to improve learning techniques. The objective of this paper is to provide an extensive analysis of current research which aims to utilize machine learning techniques to diagnose dementia and thus further enhance the advancement of more efficient method for the detection of dementia by consolidating the existing information in this sector. A systematic review was conducted utilising the PRISMA technique, which stands for Preferred Reporting Items for Systematic Reviews and Meta-Analyses. A bibliometric survey with keywords for Dementia and different machine learning techniques was carried out in the Science direct, IEEE, Springer and PubMed databases. It conducts a comprehensive examination of various Machine Learning studies. It also has a thorough analysis of Deep Learning (DL) based work. Initially, the identification stage records the number of results obtained from database searches, which amounts to 7112, screening step retrieved 2190 records and finally 65 records were included for analysis. Also, SVM and K-Nearest Neighbors (KNN) classifiers are trained on Electroencephalogram based dataset developed at Florida State University and have achieved 97.1% and 95.3% accuracy for SVM and KNN respectively. This study provides an extensive analysis of current research that aims to utilize machine learning techniques to diagnose dementia, along with a demonstration of EEG-based dementia detection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Comprehensive Analysis of Machine Learning Algorithms For Predictive Modeling in Dementia Detection

  • 期刊论文

  • 发表日期：

    2025年7月10日

查看全文
  • Hemlata Sandip Ohal,
  • Shamla Mantri

摘要

Dementia is an irreversible neurological condition that primarily impacts older individuals. The development of effective automated approaches is necessary to achieve timely and accurate detection. Various innovative methodologies have been suggested by researchers to categorize dementia. However, it is crucial to have a thorough comprehension of the current research to improve learning techniques. The objective of this paper is to provide an extensive analysis of current research which aims to utilize machine learning techniques to diagnose dementia and thus further enhance the advancement of more efficient method for the detection of dementia by consolidating the existing information in this sector. A systematic review was conducted utilising the PRISMA technique, which stands for Preferred Reporting Items for Systematic Reviews and Meta-Analyses. A bibliometric survey with keywords for Dementia and different machine learning techniques was carried out in the Science direct, IEEE, Springer and PubMed databases. It conducts a comprehensive examination of various Machine Learning studies. It also has a thorough analysis of Deep Learning (DL) based work. Initially, the identification stage records the number of results obtained from database searches, which amounts to 7112, screening step retrieved 2190 records and finally 65 records were included for analysis. Also, SVM and K-Nearest Neighbors (KNN) classifiers are trained on Electroencephalogram based dataset developed at Florida State University and have achieved 97.1% and 95.3% accuracy for SVM and KNN respectively. This study provides an extensive analysis of current research that aims to utilize machine learning techniques to diagnose dementia, along with a demonstration of EEG-based dementia detection.