An artificial neural network approach for classification of vector-borne diseases

Vector-Borne diseases are quite prevalent in India and cause a large number of deaths when they get aggravated, in turn leading to epidemics. It is quite easy to get infected by these diseases, which have very similar symptoms, most of which manifest after days. Technology, today, can provide a helping hand in the correct diagnosis of these diseases. In this paper, we take up three diseases prevalent in India: malaria, dengue and chikungunya. The proposed method uses an Artificial Neural Network (ANN) based backpropagation algorithm for training and testing. A number of gradient optimization techniques are used like Adaptive Moment Estimation, RMSProp, Adagrad, Classical Momentum and Nesterov accelerated gradient. The final probability of the most probable of the three diseases is given, based on the symptoms entered. Using backpropagation algorithm, an accuracy of 99.7% was achieved.

[1]  Tito Waluyo Purboyo,et al.  Stomach disorder detection through the Iris Image using Backpropagation Neural Network , 2016, 2016 International Conference on Informatics and Computing (ICIC).

[2]  Sankhadip Saha,et al.  Combined committee machine for classifying dengue fever , 2016, 2016 International Conference on Microelectronics, Computing and Communications (MicroCom).

[3]  Anmol Aggarwal,et al.  A novel method for medical disease diagnosis using artificial neural networks based on backpropagation algorithm , 2013 .

[4]  Xiangji Huang,et al.  CNN-based image analysis for malaria diagnosis , 2016, 2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM).

[5]  S. M. Rafiuddin,et al.  Performance analysis of supervised machine learning algorithms for text classification , 2016, 2016 19th International Conference on Computer and Information Technology (ICCIT).

[6]  Nisar Hundewale,et al.  Comparison of classification techniques-SVM and naives bayes to predict the Arboviral disease-Dengue , 2011, 2011 IEEE International Conference on Bioinformatics and Biomedicine Workshops (BIBMW).

[7]  Georg Heigold,et al.  An empirical study of learning rates in deep neural networks for speech recognition , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.