ABSTRACT
Data mining processes are used to build machine learning models that power applications including search engine technology and website recommendation programs. The aim of this thesis is to form a system and regard an operative to accomplish the investment projects objective in term of both duration and charge by using the techniques: genetic algorithm (GA) and gravitational search algorithm (GSA). To achieve the research aim of the research a system was built in order to optimize the solution for the investment projects problems and increase the number of projects to be implemented. The results show that the there is confusion in the work of the Iraqi governments and the national Investment Commission in creating a favorable investment environment that holds the elements of a successful investment environment. GSA find the objective function according to its round from 4 to 5 iteration while GA more than 8 iteration. GSA algorithm operative method in determining the greatest solution and it shorter time. GA technique consume a lot of time in finding the solution and managing the investment projects and sometime yield to non-optimal solution. As comparing to GSA, the value of the objective function for the profit that is chosen is very low and leads to reduce the number of the project implemented, as GA is fall in under fitting and the steps as selection, cross over and mutation lead to this value. The difference between GA and GSA is also very clear in the effectiveness of the solution, as xij also has lower value than GSA which lead to reduce the effectiveness. In most of the governorate the number of the projects that can be maximize increase more than its originals as they reach 91 projects as in Baghdad with profit more than 4 billion dollar. The number of project to be implemented by using GSA in National governorate about 101 projects comparing with GA which only give 51 projects.In Diyala Governorate, 5 projects were taken and each project has three significant problems with each one of them has three solutions, GSA found the best solution with effectiveness of 0.89 and the same for GA and its seem to be equal in number of problems which mean show an equal effect.
