Scholarlink

Many evolutionary changes have occurred in today’s environment and human lifestyles, causing various neurodevelopmental disorders. One of the most significant disorders among all other disorders is attention deficit hyperactivity disorder (ADHD). It has a substantial occurrence percentage. The diagnosing procedure for ADHD deals with analysing a large amount of medical data of the ADHD individual and frequently counselling him accordingly. It takes a lot of quality time for the person. It is essential to identify ADHD in its early stages for effective treatment. General statistical survey data by National Survey of Children’s Health (NSCH) is used for predicting the occurrence of ADHD in people using machine learning models. Since this survey data is multi-dimensional, certain dimensionality reduction techniques are used to extract features that cause ADHD predominantly. Feature subsets are extracted from these predominant features using feature extraction methodologies. Various machine learning algorithms are developed using these feature subsets and their respective performances are analysed to find a suitable classifier that predicts ADHD neurodevelopmental disorder. The performance metrics of Naive Bayesian and K Nearest Neighbour classifiers are enhanced as the significant features were chosen. However, the performance of the Decision Tree classifier is outstanding with a score of 91% and the lowest training time of 29.60 s.

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

Prediction of Attention Deficient Hyperactivity Disorder from Multi-dimensionality Dataset Using Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Pottem Archana,
  • Sudheer Kumar Terlapu,
  • M. Srilakshmi,
  • M. Venkata Subbarao,
  • Viswanadham Ravuri

摘要

Many evolutionary changes have occurred in today’s environment and human lifestyles, causing various neurodevelopmental disorders. One of the most significant disorders among all other disorders is attention deficit hyperactivity disorder (ADHD). It has a substantial occurrence percentage. The diagnosing procedure for ADHD deals with analysing a large amount of medical data of the ADHD individual and frequently counselling him accordingly. It takes a lot of quality time for the person. It is essential to identify ADHD in its early stages for effective treatment. General statistical survey data by National Survey of Children’s Health (NSCH) is used for predicting the occurrence of ADHD in people using machine learning models. Since this survey data is multi-dimensional, certain dimensionality reduction techniques are used to extract features that cause ADHD predominantly. Feature subsets are extracted from these predominant features using feature extraction methodologies. Various machine learning algorithms are developed using these feature subsets and their respective performances are analysed to find a suitable classifier that predicts ADHD neurodevelopmental disorder. The performance metrics of Naive Bayesian and K Nearest Neighbour classifiers are enhanced as the significant features were chosen. However, the performance of the Decision Tree classifier is outstanding with a score of 91% and the lowest training time of 29.60 s.