Optimal Decision Making for Deep Learning-Based Prediction with Combinatorial Optimization

The growing advancement of technology helps researchers a lot to confront new problems and making new solutions. Not only with the existing algorithm but also various ways can be found for making a solution. This could be either mathematically or logically. This work proposed a method for using deep learning for solving the combinatorial optimization problem which is the most talked about the topic now in the field of computer science. The combinatorial optimization problem is very much hard in nature and very much difficult to solve or compute mathematically. This proposed work help to find the optimization with the help of deep learning. This work mainly focuses on a problem that is considered as a data point and finds out the optimal solution with the help of deep learning for the given task. Such as for a problem set of mango leave disease, this proposed work at first by applying some Deep learning methods with Convolutional Neural Network models such as DenseNet201, and Inception V3. Also, the accuracy of these models is again justified by seven performance metrics like Accuracy, Precision, F1 Score, Sensitivity, Specificity, FNR, FPR. After getting the full model the optimal solution finds using the combinatorial optimization algorithm. Which gives us the best solution between the two models.