A bidding strategy based on genetic network programming in continuous double auctions

Nowadays, along with the increasing convenience of internet, there becomes popular electronic type of doubleside auctions, continuous double auction (CDA), where multiple sellers and multiple buyers can update their asks and bids continuously at the same time through the trading period. In this paper, a bidding strategy using Genetic Network Programming (GNP) with Rectifying Node has been proposed to deal with CDA automatically based on our previous works of English auction and Dutch auction. The stategy considers various aspects of environments to determine the most competitive and suitable bid price at each time step without losing profit. The agent can manage complex CDA situations more effectively and efficiently by employing the GNP structure using the auction information and heuristic control functions. Experiments were conducted on a series of CDAs under different settings, and compared with some other methods.