Date of Award

December 2016

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering and Computer Science


Kishan G. Mehrotra

Second Advisor

Chilukuri K. Mohan


Key Players, Multi-objective optimization, Networks

Subject Categories



Identification of a set of key players, is of interest in many disciplines such as sociology, politics, finance, economics, etc. Although many algorithms have been proposed to identify a set of key players, each emphasizes a single objective of interest. Consequently, the prevailing deficiency of each of these methods is that, they perform well only when we consider their objective of interest as the only characteristic that the set of key players should have. But in complicated real life applications, we need a set of key players which can perform well with respect to multiple objectives of interest.

In this dissertation, a new perspective for key player identification is proposed, based on optimizing multiple objectives of interest. The proposed approach is useful in identifying both key nodes and key edges in networks. Experimental results show that the sets of key players which optimize multiple objectives perform better than the key players identified using existing algorithms, in multiple applications such as eventual influence limitation problem, immunization problem, improving the fault tolerance of the smart grid, etc.

We utilize multi-objective optimization algorithms to optimize a set of objectives for a particular application. A large number of solutions are obtained when the number of objectives is high and the objectives are uncorrelated. But decision-makers usually require one or two solutions for their applications. In addition, the computational time required for multi-objective optimization increases with the number of objectives. A novel approach to obtain a subset of the Pareto optimal solutions is proposed and shown to alleviate the aforementioned problems.

As the size and the complexity of the networks increase, so does the computational effort needed to compute the network analysis measures. We show that degree centrality based network sampling can be used to reduce the running times without compromising the quality of key nodes obtained.


Open Access

Included in

Engineering Commons