Lumbar Spine Disease Prediction with KNN, Random Forest and Decision Tree: A Study

This research provides the in-depth analysis of mechanisms used to predict Lumbar Spine related diseases. Machine learning based mechanisms are most common towards prediction of Lumbar Spine related diseases. The empirical study consists of phases including pre-processing, feature extraction, training, and testing. Pre-processing used to remove the noise from the dataset. The noise handling mechanisms includes median filtering and Gaussian filtering. The feature extraction mechanisms are employed to extract the attributes necessary for the training phase. The training phase performed to make the model learn about the symptoms and diseases present within the dataset. The last phase consists of classification. The classification phase generally consists of KNN, Random Forest and decision tree. Validation is accomplished with the help of metrics including classification accuracy, specificity, sensitivity, and F-Score. These metrics must be improved to generate the accurate result about the disease.