Future Trails for Integer Programming and Relations to Artificial Intelligence

Aims: A review of methods and approaches for solving linear integer problems is presented in this work. These problems are classified as NP-hard optimization algorithms in artificial intelligence.

Study Design: we have used the Google scholar to collect the data resources from past 5 years to analysis the techniques and methods used in different algorithms in artificial intelligence.

Methodology: Exact optimum solution for this class of challenges also need use of substantial computer resources. The current direction in which several researcher focuses their efforts to effectively address numerous difficult practical issues is the creation of efficient hybrid techniques that combine in an appropriate way the finest elements of multiple methods (precise or estimated). The approximation algorithms' core heuristic techniques might be classified as constructive algorithms and local-improvement algorithms.

Results: We examined three artificial intelligence algorithms utilizing the linear integer programming approach. Algorithm based on population It has also been demonstrated that a population of a critical size is necessary for a population-based optimization method to be effective. The genetic algorithm is shown next. The goal value associated with this solution may be utilized to effectively reduce the search tree in bound and branch type integer programming methods. Finally, we analyze the particle swarm optimization (PSO) approach, which demonstrates that In most cases, PSO outperforms the Branch and Bound method in solving such issues quickly.

Conclusion: In actuality, integer optimization issues describe a wide spectrum of real-world difficulties. Their population and size are constantly growing. Although while accurate methods for integer issues have substantially improved in recent years, their long runtimes and memory needs make them unsuitable for actual medium and large-scale applications.