Worked on the NP-Hard problem of Combinatorial Auction i.e. maximisation of revenue collection. Modelled the problem as a local search and constraint satisfaction problem and used greedy hill climbing with random restarts and random walks.
The algorithm uses Local search with random restarts for finding better allocation of the available resource. The Problem Statement can be seen in the pdf.