Performance Enhancement to Improve Accuracy for the Identification of Lyme Disease by using Novel ANN Algorithm by Comparing with K-Mean Algorithm
M. Saravanan, Shail Shakti
Abstract
The goal of this study is to evaluate the accuracy of novel Artificial Neural Network (ANN) machine learning models to the K-mean method in identifying Lyme disease in humans. Materials and Methods: The two groups in the research are the novel ANN machine learning model and the K-mean method. The dataset utilized in this study is a sample set of 889 photos from Kaggle that was used to identify Lyme infection. Training and test datasets were built from the dataset. With a G power of 80%, accuracy was estimated using a training dataset of 716 images and a test dataset of 173 images. The experiment was iterated 20 times using the mentioned models. Results and Discussion: For the dataset on Lyme disease, it was shown that the novel ANN model achieved an accuracy of 89% compared to the K-mean algorithm’s 87%. It is obvious that the novel Artificial Neural Network is shown to have superior statistical significance than the K-mean model, with a p value of 0.001, which is less than 0.05 in the independent t-test. The findings indicate that the novel Artificial Neural Network approach is beneficial for early Lyme disease prediction and successful at recognizing skin lesions.