Analysis of clinical texts for prediction of COVID-19 using Bag of Words and Artificial Neural Networks
Main Article Content
Abstract
In recent times, the number of COVID-19 cases worldwide has increased significantly, which led to overwhelming work for healthcare workers and forced many countries to go under complete lockdown. Thus, it is essential to find ways to detect and control coronavirus. Machine learning has shown promising results in the various medical fields by analyzing clinical data, so it is crucial to explore new ways to detect COVID-19. It is challenging to test everyone, so a controlled and automated system to detect the COVID-19 is needed nowadays. These days, an enormous amount of data related to COVID-19 is available. In this work, we propose an artificial neural network and bag of words model-based approach for classifying clinical reports of patients into four classes of the virus. The features were extracted using various techniques like Term frequency/Inverse document frequency (TF/IDF) and report length, which is then passed through a robust neural network classifier. We trained and tested different neural network models to find suitable architecture. The model showed better performance than all the classical and ensemble machine learning algorithms with an accuracy of 97.2%. It offers excellent potential for the early detection of COVID-19 and thus helps control the pandemic.
Downloads
Metrics
Article Details
Licensing
TURCOMAT publishes articles under the Creative Commons Attribution 4.0 International License (CC BY 4.0). This licensing allows for any use of the work, provided the original author(s) and source are credited, thereby facilitating the free exchange and use of research for the advancement of knowledge.
Detailed Licensing Terms
Attribution (BY): Users must give appropriate credit, provide a link to the license, and indicate if changes were made. Users may do so in any reasonable manner, but not in any way that suggests the licensor endorses them or their use.
No Additional Restrictions: Users may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.