A Comparative Analysis of Machine Learning Classifiers for Audio-Based Disease Classification
B. Varun, Ishita Jonnalagadda, Ch Vishnu Vardhan, Marimganti Srinivas, Sahithya Ullas · 2024
In order to address the increasingly significant role played by the monitoring of health issues during the times of growing healthcare challenges, this research paper will discuss more about the use of audio signals for diagnosing respiratory diseases by the means of analyzing of cough sounds, breathing patterns, and other audio cues. However, the vision of this work comes from the desire to make a non-invasive and efficient diagnostic process to overcome the difficulty of the traditional invasive ways of imageries. Combining the machine learning techniques using Mel-Frequency Cepstral Coefficients (MFCC) for feature extraction we aimed to construct a powerful tool that can do health state labels using the audio signals with accuracy. In the end, our multi-classifier approach in conjunction with accuracy measurements reveal the promising applications of machine learning algorithms to accurately differentiate between various respiratory diseases, thereby opening the avenue for more robust health monitoring systems.