Machine Learning Approach to Predict Asthma Prevalence with Decision Trees

Abstract

Deep Learning and Machine Learning algorithms are popularly used in the healthcare sector for diagnosis with many algorithms have been successfully implemented to perform patient disease diagnosis and treatment plans. Decision Tree algorithms are profoundly used in the healthcare industry to implement the methods for various disease diagnoses, predictions, therapeutic recommendations, automated tasks, and communication between patients and customer service. Decision Trees work effectively with classification as well as regression techniques. Decision Trees are easy and swift to efficiently implement for faster outcomes in disease diagnosis and they are comprehensively used in data mining and decision-making processes. Decision Trees conjoined with ensemble methods such as Random Forest and Gradient Boost, enhance the performance and accuracy of results in predictions associated with regression tasks. Asthma is an inflammatory and chronic disease that affects a large population worldwide, with severe conditions resulting in emergency visits to the hospital. Asthma is a lung disease caused by airway inflammation and the airways become sensitive to allergic substances. Timely detection of this disease wards off undesirable events, and critical care visits, and is the basis for a good prognosis for patient recovery. Precautionary measures possibly reduce the complications of disease progression by knowing the disease level and associated complications at an early stage. This research article wants to focus on the best model for predicting Asthma prevalence with Decision Tree algorithms as these techniques work faster and provide quicker reports. The Weka tool was used for the model creation with datasets downloaded from data.world and The California Department of Public Health's Open Data Portal.

Topics & Concepts