Automated Bidding Strategy using Genetic Algorithm for Online Auctions

Due to the proliferation of online auctions, there is an increasing need to monitor and bid in multiple auctions in order to procure the best deal for the desired good. This paper reports on the development of a heuristic decision making framework that an agent can exploit to tackle the problem of bidding across multiple auctions with varying start and end times and with varying protocols. As the range of potential strategies is huge, we decided to use a genetic algorithm (GA) to search for effective strategies for each of the various environments that we identified. This strategy is termed the intelligent bidding strategy in the remainder of this article. Finally, we systematically evaluate the intelligent bidding strategy to highlight its operational characteristics in different scenarios and present our conclusions and further work.