Mangesh Gharote

Stable Matching (SM) has received a lot of attention from researchers due to its useful applications in practice. Gale and Shapley were the first to propose a polynomial-time algorithm to find an SM solution, for matching with strict preferences. However, their algorithm often produces extreme stable solutions, favouring either men or women. In practice, the real-life problems have multiple objectives (equitable and welfare) and preferences are not strict, consequently ties occur. With the inclusion of ties and objectives, the problem becomes NP-hard. A few researchers proposed local search and evolutionary algorithms to solve multi-objective SM problem with ties, but these methods were not scalable. In this paper, we propose an efficient Goal Programming and Repair Heuristics based approach to solve this problem. On comparison with other related works, our approach shows significant improvement in respective objectives (equity and welfare). This approach with slight modification can be proved useful for solving other hard variants of the SM problem.

In this paper, we address the problem of allocating workforce to software projects, considering bilateral (project managers' and employees') preferences and cost of allocation. Earlier research has addressed the workforce allocation to software projects as stable matching problem with strict preferences ranking. However, in practice, there are scenarios where project managers did not strictly rank the employees; therefore, there are ties in the preference list. We developed a stable matching with ties-based optimization model to create a minimum cost stable allocation. Our results suggest that a software firm can reduce allocation cost and improve the average preferences for both the workforce and the project managers in the presence of ties in the preference list.

In this paper, we compare organizational units using Data Envelopment Analysis (DEA) on the basis of Human Capital (HC) measures. HC based performance measurement has become an important aspect for organizations to sustain long-term growth. With the inclusion of HC measures, the variable selection for DEAmodel becomes challenging. We propose an automated model discovery method for variable selection. The guideline for variable selection is based on the information obtained from the executives about the variables. This information about the variables could be complete, partial or incomplete. We demonstrate our method under these various scenarios. The methodology is applied to compare 40 units of an IT service firm. Further, we discuss in detail the managerial insights for improvement of the inefficient units.