Comparison of accuracy for conditional random field algorithm and multilayer perceptron in pathological voice detection

The point of this study is to create and analyze the exhibition of two different AI calculations for the discovery of obsessive voices on medical services frameworks and virtual entertainment applications. The dataset utilized in this paper is the Pathology Voice Identification Values dataset, which incorporates 41,090 preparation voice documents. In this review, the dataset is alluded to as the Voice Discovery dataset. The information source connection can be gotten to from ("Kaggle" 2022) site. Two gatherings were utilized in this article for execution. Bunch 1 is the Restrictive Irregular Field calculation and Gathering 2 is the Multi-facet Perceptron calculation. The example size is determined utilizing G power 80% for two gatherings and there are 20 examples utilized in this work for every one of the calculations. The CRF acquired an exactness of (80.50%) when contrasted with Multi-facet Perceptron with a precision of (71.26%). The importance esteem between the two gatherings is p=0.000 (Autonomous Example T-test p<0.05). The CRF calculation acquired higher exactness with 80.50% when contrasted with the Multi-facet Perceptron calculation with 71.26%.