Consolidation of the IFM with the JSSP through Neural Networks as Model for Software Projects

This paper proposes a consolidation of the Incremental Funding Method (IFM) with the Job Shop Scheduling Problem (JSSP) through Neural Networks, in order to develop a model for software projects. More specifically, it formulates the IFM method in terms of JSSP model and then gives a solution method based on Recurrent Neural Networks (RRNs). The IFM is a financial approach to software project management aiming at maximizing the net present value (NPV) and it can be used in cooperation with other software development processes. However, there are few proposed algorithms that focus on the IFM modeled problem´s solving with limited results. Our goal is to employ the JSSP, a well established model, in order to model a IFM problem as JSSP problem and to propose a Neural Network solving method that gives promising results.