Journal Article

The classification of medical and botanical data through majority voting using artificial neural network

Khan, Fayaz Ahmed, Khanday, Akib Mohi Ud Din, Tripathi, Kshitij, Nisa, Khair U. L.

2023

Overview

Data classification has many approaches in data mining and machine learning. The artificial neural network (ANN) is applied to classify the data that might belong to various domains like chemical, botanical, medical, spatial, textual, and image. In this work, an ANN technique is applied to the 7 Life sciences (botanical and medical) data sets extracted from public data repositories. Various optimization approaches like exhaustive validation, cross-validation, and multiple seeding are used to discover the most optimized networks for the given datasets. Finally, voting predicts the class where the whole dataset is used as a test set instead of folds. The results obtained by the proposed approach outperform other approaches on all the datasets. Cleveland’s heart, Statlog heart, Dermatology, Hepatitis, Seeds, Abalone and Vertebral Column data sets (all of UCI) after applying the voting showed the accuracy of 94.61%, 93.7%, 99.73%, 96.77%, 99.05%, 89.37% and 90.32% respectively. In the future deep neural network may be used to improve the results.

• Publisher: Springer Nature Singapore,Springer Nature B.V
• Subject: Accuracy / Algorithms / Artificial Intelligence / Artificial neural networks / Cardiovascular disease / Classification / Computer Imaging / Computer Science / Data analysis / Data mining / Datasets / Decision trees / Dermatology / Feature selection / Heart / Image Processing and Computer Vision / Literature reviews / Machine Learning / Medical research / Neural networks / Optimization / Original Research / Pattern Recognition and Graphics / Regression analysis / Seeds / Software Engineering / Support vector machines / Vision