USING NEURAL NETWORK AND FRACTALS FRACTIONAL ANALYSIS TO PREDICT THE EYE DISEASE INFECTION CAUSED BY CONJUNCTIVITIS VIRUS

Abstract

The fractals and fractions differential operations along with artificial intelligence-based neural networks are used to study eye infectious disease caused by conjunctivitis virus. The deduction of pertinent results for the existence theory of model is reported and analyzed by using the fixed point theory. Moreover, the stability results related to Hyers–Ulam type are also investigated. Basic reproductive number together with equilibrium points is derived. Results related to sensitivity of basic reproductive number are debated. In recent times, machine learning tools have been attracting attention very well, and the artificial neural networks and deep neural networks techniques are increasingly used to investigate epidemic models. Therefore, for developing the artificial neural networking (ANN) model, a sample of 186 infected people is collected by distributing 130, 28, and 28 bandwidth for training, testing, and validation. Both Sigmoid and pureline transfer functions are used in layers. Liveners–Marquardt backward propagation algorithm (LMBA) is used for training of neural model. The constructed model is validated by conducting regression analysis. The outcomes, namely, coefficient of determination and mean square error suggest that the present ANN model is the most effective one which predicts the spread of eye diseases and can be a useful tool for controlling hazardous diseases.