The Use of Artificial Neural Networks for Estimation of Decision Surfaces in First Price Sealed Bid Auctions
The Use of Artificial Neural Networks for Estimation of Decision Surfaces in First Price Sealed Bid Auctions
复制标题
使用人工神经网络估计首价密封投标拍卖中的决策面
DOI：
10.1007/978-94-011-0770-9_2
复制
发表时间：
1994
期刊：
影响因子：
--
通讯作者：
M. Boening
中科院分区：
文献类型：
--
作者：
R. Dorsey;John D. Johnson;M. Boening
ABSTRACf. Artificial neural networks, optimized using genetic algorithms, are used to estimate bid functions for first price sealed bid auctions. Data generated in experimental markets is used for two means of estimating the bid function. First, the neural network provides a best fit to the data, thus estimating the bid function that subjects were using. Alternative objective functions are used for the neural network to demonstrate the effect on the resultant bid function. Second, the neural network is optimized using profit maximization as the objective function to identify the optimal bid function given the bids of the experimental subjects. l. INTRODUCflONArtificial neural networks have qualities that in many ways parallel human decision processes and thus may provide insight into people's behavior in specific economic environments." Neural nets" can learn associative patterns and approximate the functional relationship between the information provided to economic agents (the inputs) and agents' responses (the outputs). Much like nonlinear regression, the neural net can estimate complex decision surfaces. However, it does not require a pre-specified model and it can achieve a continuous approximation to any functional mapping between inputs and outputs. The neural network arrangement of simple nodes into multiple layers connected through nonlinear functions produces a mapping consistent with any underlying functional relationship, regardless of its complexity even in the presence of noise, distortion, and missing data. In this study, neural networks, optimized using the genetic adaptive neural network algorithm (GANNT) of Dorsey et al.(1992b), are used to estimate bid functions of bidders in first price sealed bid auctions. The point of departure is the experimental markets reported in Cox et ai.(1988). Although consistent patterns of behavior emerged in those markets, the behavior was sometimes at odds with the prevailing Nash eqUilibrium bidding theory. We
ABSTRACf. Artificial neural networks, optimized using genetic algorithms, are used to estimate bid functions for first price sealed bid auctions. Data generated in experimental markets is used for two means of estimating the bid function. First, the neural network provides a best fit to the data, thus estimating the bid function that subjects were using. Alternative objective functions are used for the neural network to demonstrate the effect on the resultant bid function. Second, the neural network is optimized using profit maximization as the objective function to identify the optimal bid function given the bids of the experimental subjects. l. INTRODUCflONArtificial neural networks have qualities that in many ways parallel human decision processes and thus may provide insight into people's behavior in specific economic environments." Neural nets" can learn associative patterns and approximate the functional relationship between the information provided to economic agents (the inputs) and agents' responses (the outputs). Much like nonlinear regression, the neural net can estimate complex decision surfaces. However, it does not require a pre-specified model and it can achieve a continuous approximation to any functional mapping between inputs and outputs. The neural network arrangement of simple nodes into multiple layers connected through nonlinear functions produces a mapping consistent with any underlying functional relationship, regardless of its complexity even in the presence of noise, distortion, and missing data. In this study, neural networks, optimized using the genetic adaptive neural network algorithm (GANNT) of Dorsey et al.(1992b), are used to estimate bid functions of bidders in first price sealed bid auctions. The point of departure is the experimental markets reported in Cox et ai.(1988). Although consistent patterns of behavior emerged in those markets, the behavior was sometimes at odds with the prevailing Nash eqUilibrium bidding theory. We