Comparative Analysis of Classifiers for Prediction of Epileptic Seizures

Epilepsy is a neurological disease in which people suffer from seizure attack and lose the normal function of brain. Almost 50 million people have epilepsy in the world due to which it has become the most common neurological disease. Early prediction of epilepsy helps patients to avoid epilepsy and live normal life. Many studies have been conducted for the early prediction of epilepsy. However, selection of the most appropriate classifier has always been a question that needs to be resolved. In this study, we are using six classifiers of machine learning which are KNN, Naïve Bayes, Linear Classification Model, Discriminant Analysis Model, Support Vector Machine and Decision Tree, to find the best classifier for the prediction of epileptic seizures, in term of accuracy. Dataset from “Kaggle” was used. Preprocessing and cross-validation of the data was carried out for training and testing of classifiers. The results depict that Naive Bayes classifier has a better average accuracy of 95.739% as compared to other classifiers. The future work of this study is to implement the suggested model in real time, so that the workload of medical members could be reduced.

Higher-Order Techniques for Heat Equation Subject to Non-Local Specifications

Higher-order numerical techniques are developed for the solution of (i) homogeneous heat equation u t = u xx and (ii) inhomogeneous heat equation u t = u xx + s(x, t) subject to initial condition u(x, 0) = f (x), 0 < x < 1, boundary condition u(0, t) = g(t)0 < t ≤ T and with non-local boundary condition(s) (i) b 0 u(x, t)dx = M (t) 0 < t ≤ T, 0 < b < 1 (ii) u(0, t) = (iii) u(1, t) = 1 0 φ(x, t)u(x, t)dx + g 1 (t), 0 < t ≤ T and 1 0 ψ(x, t)u(x, t)dx + g 2 (t), 0 < t ≤ T as appropriate. The integral conditions are approximated using Simpson’s 1 3 rule while the space derivatives are approximated by higher-order finite difference approxi- mations. Then method of lines, semidiscritization approach, is used to trans- form the model partial differential equations into systems of first-order linear ordinary differential equations whose solutions satisfy recurrence relations in- volving matrix exponential functions. The methods are higher-order accurate in space and time and do not require the use of complex arithmetic. Parallel algorithms are also developed and implemented on several problems from lit- erature and are found to be highly accurate. Solutions of these problems are compared with the exact solutions and the solutions obtained by alternative techniques where available.