Piecewise Pseudo-Maximum Likelihood Estimation for Risk Aversion Case in First-Price Sealed-Bid Auction

In the literature, more and more clues show that the bidders indeed tend to risk averse. But as a good structure estimation procedure in first-price sealed-bid auctions, the traditional PPMLE (Piecewise Pseudo-Maximum Likelihood Estimation) approach is only applicable to the risk neutrality case. This study generalizes this approach to the risk aversion case, and obtains a simple estimation procedure for the Pareto family of private costs. In the procedure, the estimation order of distribution parameters of private costs is not important, which is different from the traditional PPMLE method. Though there is no unique estimator for θ ₁ and γ, some guidance is given for real-world applications on the basis of our Monte Carlo simulation experiments. What’s more, this procedure can be reduced further, if one is only interested in estimating the private costs. Extensive simulation experimental results indicate that our approach outperforms or matches at least the traditional one in the case of risk neutrality, and is applicable to the risk aversion case.     

Structure damage diagnosis using neural network and feature fusion

A structure damage diagnosis method combining the wavelet packet decomposition, multi-sensor feature fusion theory and neural network pattern classification was presented. Firstly, vibration signals gathered from sensors were decomposed using orthogonal wavelet. Secondly, the relative energy of decomposed frequency band was calculated. Thirdly, the input feature vectors of neural network classifier were built by fusing wavelet packet relative energy distribution of these sensors. Finally, with the trained classifier, damage diagnosis and assessment was realized. The result indicates that, a much more precise and reliable diagnosis information is obtained and the diagnosis accuracy is improved as well.     

Performance modeling,real-time dispatching and simulation of wafer fabrication systems using timed extended object-oriented Petri nets

Semiconductor wafer fabrication system (SWFS) is the most complex and capital-intensive phase in the entire semiconductor manufacturing cycle. With characteristics of re-entrant processing routes, equipment uncertainty, product diversity and improving technologies, great challenges are presented in SWFS’s modeling, scheduling and simulation. To implement efficient production control, this paper provides a timed extended object-oriented Petri nets (TEOPNs) approach to performance modeling, real-timed dispatching and simulation of SWFSs. The TEOPNs models are constructed in a hierarchy to accord with the real-world SWFS’s organization, and a new type of signal place is added into the TEOPNs to respond the dynamic states of all processing facilities. A novel autonomy and coordination-based real-time dispatching mechanism (A&C-RDM) is developed in this paper, which executes under the support of the TEOPNs-based hybrid real-time dispatching control system (HRDCS). Owning to the ability of gathering dynamic real-time information of all production facilities and WIP products, the HRDCS can make adaptive dispatching decisions according to the local and global real-time processing status. Two sets of key elements of real-time dispatching, i.e. the state thresholds and dispatching rules, are defined in the HRDCS so that the A&C-RDM can integrate different types of dispatching rules. A set of simulation experiments prove the efficiency of the proposed modeling and dispatching algorithm. In summary, the proposed TEOPNs, HRDCS and A&C-RDM form the cornerstones of a real-time dispatching simulation prototype of SWFS, and the work described in this paper carries out an advanced integrated “modeling–dispatching–simulation” methodology.     

Parameterized computational complexity of control problems in voting systems

Voting systems are common tools in a variety of areas. This paper studies parameterized computational complexity of control of Plurality, Condorcet and Approval voting systems, respectively. The types of controls considered include adding or deleting candidates or voters, under constructive or destructive setting. We obtain the following results: (1) constructive control by adding candidates in Plurality voting is W[2]-hard with respect to the parameter “number of added candidates”, (2) destructive control by adding candidates in Plurality voting is W[2]-hard with respect to the parameter “number of added candidates”, (3) constructive control by adding voters in Condorcet voting is W[1]-hard with respect to the parameter “number of added voters”, (4) constructive control by deleting voters in Condorcet voting is W[1]-hard with respect to the parameter “number of deleted voters”, (5) constructive control by adding voters in Approval voting is W[1]-hard with respect to the parameter “number of added voters”, and (6) constructive control by deleting voters in Approval voting is W[2]-hard with respect to the parameter “number of deleted voters”.     

An expert system to classify microarray gene expression data using gene selection by decision tree

Gene selection can help the analysis of microarray gene expression data. However, it is very difficult to obtain a satisfactory classification result by machine learning techniques because of both the curse-of-dimensionality problem and the over-fitting problem. That is, the dimensions of the features are too large but the samples are too few. In this study, we designed an approach that attempts to avoid these two problems and then used it to select a small set of significant biomarker genes for diagnosis. Finally, we attempted to use these markers for the classification of cancer. This approach was tested the approach on a number of microarray datasets in order to demonstrate that it performs well and is both useful and reliable.