An Enriched Intuitionistic Kernel Based K-Medoids Clustering for Indeterminacy Handling in ADHD Prediction
M.Lalithambigai1, A.Hema2

1M.Lalithambigai*, Computer Science,Kongunadu Arts and Science College.
2Dr.A.Hema, Computer Science,Kongunadu Arts and Science College.

Manuscript received on 2 August 2019. | Revised Manuscript received on 10 August 2019. | Manuscript published on 30 September 2019. | PP: 2815-2820 | Volume-8 Issue-3 September 2019 | Retrieval Number: C5207098319/2019©BEIESP | DOI: 10.35940/ijrte.C5207.098319
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: In recent year it is revealed that prevalence of attention-deficit/hyperactivity disorder (ADHD) among primary school children’s is widespread. ADHD is considered as one of the most common childhood disorders and can endure through adolescence and adulthood. Addressing and accurate diagnosis of ADHD in earlier stages will be very effective for proper and timely treatment. But it is very complex to differentiate behaviour that reflect ADHD victim from the normal growth. Though there are several existing works are available for detecting ADHD using machine learning handling indeterminacy is a toughest challenge among researchers. This paper aims at developing an unsupervised learning model-based feature subset selection to eradicate the problem of indeterminacy in handling ADHD prediction. This work adapted introduced the concept of intuitionistic kernel-based k-medoids clustering (IKKMC) for grouping similar type of ADHD patients through the knowledge of degree of membership and degree of hesitation. In this work the outliers are easily handled with intuitionistic fuzzy logic. After performing clustering, the potential feature subset involved in ADHD prediction is identified by applying Recursive Feature elimination model. The simulation results provide the evidence for IKKMC with RFE selected feature subset increases the prediction process of ADHD more accurately than other state of art.
Keywords: Attention-deficit/hyperactivity Disorder, Intuitionistic Fuzzy, Recursive Feature Elimination, Kernel, K-Medoids and Indeterminacy.

Scope of the Article:
Clustering