A comparative study of machine learning techniques for accurate disease prediction using symptom-based diagnosis

Several important data mining techniques have been developed and used in real-world settings such as in healthcare, pharmaceutical and bio-technology. This has led to the use of these techniques in conjunction with machine-learning to extract valuable information from specific data in healthcare, pharmaceutical and bio-technological sectors. Accurate predictive data analysis from the healthcare and pharmaceutical databases can help diagnose diseases promptly for treating patients and for providing services for the community. Accurate data analysis from the these database can support early disease detection, patient treatment, and community services. Like many other fields, machine learning successfully predicts these diseases. The goal of making classifier systems with Artificial Intelligence techniques to help doctors predict and diagnose diseases in their initial stages would be a big step toward solving health problems. This research highlights the comparative analysis of machine-learning algorithms like the Random Forest, Decision Tree, K-Nearest Neighbor, Naive Bayes, Support Vector Classifier, and a Deep Learning algorithm i.e., 1-Dimensional Convolutional Neural Network for an illness prediction system. A Graphical User Interface (GUI) will show the predicted results based on the best working algorithms.