Wuyi Yue

Support vector machine based multiagent ensemble learning for credit risk evaluation

In this paper, a four-stage support vector machine (SVM) based multiagent ensemble learning approach is proposed for credit risk evaluation. In the first stage, the initial dataset is divided into two independent subsets: training subset (in-sample data) and testing subset (out-of-sample data) for training and verification purposes. In the second stage, different SVM learning paradigms with much dissimilarity are constructed as intelligent agents for credit risk evaluation. In the third stage, multiple individual SVM agents are trained using training subsets and the corresponding evaluation results are also obtained. In the final stage, all individual results produced by multiple SVM agents in the previous stage are aggregated into an ensemble result. In particular, the impact of the diversity of individual intelligent agents on the generalization performance of the SVM-based multiagent ensemble learning system is examined and analyzed. For illustration, one corporate credit card application approval dataset is used to verify the effectiveness of the SVM-based multiagent ensemble learning system.

Analytical methods to calculate the performance of a cellular mobile radio communication system with hybrid channel assignment

Two approximate techniques are presented to evaluate the performance of large-scale mobile radio systems using a hybrid channel assignment scheme and a cellular structure, The two approximate analyses give the steady-state probability distributions of the system which are used to obtain expressions for the blocking probabilities. In the first method, the blocking probability is obtained by finding the interarrival time probability distribution function of one composite interrupted Poisson process (IPP) stream consisting of several IPP streams overflowing from the cell of interest and its cochannel interference cells. The second method is proposed to solve the blocking probability of the system by regarding each call as a GI/M/m(m) model. Analytical results are compared with simulation results, and good agreement is observed for both channel assignments (hybrid and fixed). The methods presented are applicable to the design of hybrid channel assignment schemes and dynamic channel assignment schemes. >

The effect of capture on performance of multichannel slotted ALOHA systems

The author presents an exact analysis to evaluate the effect of capture on the multichannel slotted ALOHA protocol. The author derives the probabilities of the successful transmission. These probabilities are used to calculate the throughputs, average packet delays for both IFT (immediate-first-transmission) and delayed-first-transmission protocols and numerically compare the performance of the systems with and without capture. Numerical results show that when a quantitative capture restriction u is considered, in a multichannel system having a fixed total bandwidth, depending on parameter u and channel number M, an improved system performance such as the average channel utilization and average packet delay can be obtained. >

Advances in queueing theory and network applications

Queueing Processes.- Two Sided DQBD Process and Solutions to the Tail Decay Rate Problem and Their Applications to the Generalized Join Shortest Queue.- Analytical Model of On-Demand Streaming Services Based on Renewal Reward Theory.- Single-Server Queues.- A Pure Decrement Service Geom/G/1 Queue with Multiple Adaptive Vacations.- Performance Analysis of an M/M/1 Working Vacation Queue with Setup Times.- Modeling of Production System with Nonrenewal Batch Input, Early Setup, and Extra Jobs.- Performance Analysis of an M/E k /1 Queue with Balking and Two Service Rates Based on a Single Vacation Policy.- Multiple Queues.- Markovian Polling Systems: Functional Computation for Mean Waiting Times and its Computational Complexity.- Performance Analysis of a Two-Station MTO/MTS Production System.- Finite-Buffer Queues.- Analysis of an M/M/c/N Queueing System with Balking, Reneging, and Synchronous Vacations.- Analysis of Mixed Loss-Delay M/M/m/K Queueing Systems with State-Dependent Arrival Rates.- Asymptotic Behavior of Loss Rate for Feedback Finite Fluid Queue with Downward Jumps.- Explicit Probability Density Function for the Length of a Busy Period in an M/M/1/K Queue.- Network Applications.- Performance Analysis of ARQ Schemes in Self-Similar Traffic.- Modeling of P2P File Sharing with a Level-Dependent QBD Process.- Performance Analysis of a Decentralized Content Delivery System with FEC Recovery.- Blocking Probabilities of Multiple Classes in IP Networks with QoS Routing.

Risk analysis of a pay to delay capacity reservation contract

This paper addresses a topic of supply chain management from a new perspective: incorporating risk aversion in a supply contract model. A pay to delay capacity reservation contract is analyzed by using the concept of conditional value-at-risk (Cvar). First, we construct the manufacturer’s optimal ordering problem by using the dynamic programming approach. Then, we derive the manufacturer’s optimal ordering strategy and compare our results with those obtained by using expectation performance and mean-variance tradeoff. Finally, numerical results are shown to reveal the impact of risk aversion on the manufacturer’s optimal decisions. The analysis presented in this paper reveals advantages of using the Cvar approach over the mean-variance approach.

Performance Analysis for a System of Connection Oriented Internet Service with a Release Delay

In this paper, we propose the use of a discrete-time connection oriented Internet service system with a release delay for broadband, high-speed, high-capacity and high-reliability Internet requirements. The release delay called close-delay is set before the release process of a connection. An upper limit length T called timer length is set as a system parameter for the close-delay period. We build a batch arrival Geom*/G/1 queue model with a setup/close-delay/close-down strategy to characterize the system operation. By using a discrete-time imbedded Markov chain approach, we derive the stationary distribution of the system, and present the formula for Probability Generation Functions of the queue length, waiting time, busy period and busy cycle. Correspondingly, we describe the performance measures for the packet response time, setup ratio, and utility of connection. We also develop a cost model to determine the optimal timer length and its expected optimal cost. Based on numerical results, we discuss the influence of the timer length for the close-delay period on the system performance and investigate the minimum timer length and the minimum cost for different offered loads and different burst degrees, and show that the choice of the timer length is significant in improving the system performance.

An exact analysis for CSMA/CA protocol in integrated voice/data wireless LANs

In this paper, we present an exact analysis for the performance of wireless LANs which support real-time and data communication using a carrier sense multiple access with collision avoidance (CSMA/CA) protocol. In the system, a user with a packet ready to transmit initially sends some pulse signals with random intervals within a collision avoidance period before transmitting the packet to verify a clear channel. The system consists of a finite number of users to share a common channel. Each user can be a source of both voice and data traffic. The time axis is slotted, and a time frame has an arbitrary large number of slots dependent on the chosen durations of the collision avoidance period and packet transmission period. A discrete-time Markov process is used to model the system operation. Numerical results are shown in terms of channel utilization and average packet delay for various packet generation rates. They indicate that the network performance can be improved by adequate choice of ratios between collision avoidance periods and transmission periods, and pulse transmission probability.

A multi-subsystem fuzzy DEA model with its application in mutual funds management companies’ competence evaluation

Abstract This paper proposes a multi-subsystem fuzzy data envelopment analysis (MFDEA) model to evaluate mutual funds management companies’ core competence which is apparently characterized with both qualitative factors and quantitative factors. As the first attempt of core competence analysis of mutual funds management companies in a financial market, we evaluate how close the mutual funds management companies are to the best practices frontier, and analyze how they should manage their funds. An empirical study of 32 mutual funds management companies in China leads to a meaningful analysis and some management insights. The approach can also be applied to many other hard assessment and evaluation problems.

Optimal replacement of a system according toa semi-Markov decision process in a semi-Markov environment

This paper investigates an optimal replacement problem of a system in a semi-Markov environment. The system itself deteriorates according to a semi-Markov process, and is further influenced by its environment, which changes according to a semi-Markov process. Each change of the environments state will change the parameters modelling the system and also cause damage on the system. For minimizing the discounted total costs with finite and infinite horizons, we show the existence of optimal control limit policies. A special case of Markov environment is discussed, and the state space can be simplified equivalently to be finite, so the real computation of the problem is feasible. Finally, a numerical example is given to prove the correctness and validity of the analysis.

Output and delay process analysis for slotted CDMA wireless communication networks with integrated voice/data transmission

This paper presents the output and delay process analysis of integrated voice/data slotted code division multiple access (CDMA) network systems with random access protocol for packet radio communications. The system model consists of a finite number of users, and each user can be a source of both voice traffic and data traffic. The allocation of codes to voice calls is given priority over that to data packets, while an admission control, which restricts the maximum number of codes available to voice sources, is considered for voice traffic so as not to monopolize the resource. Such codes allocated exclusively to voice calls are called voice codes. In addition, the system monitoring can distinguish between silent and talkspurt periods of voice sources, so that users with data packets can use the voice codes for transmission if the voice sources are silent. A discrete-time Markov process is used to model the system operation, and an exact analysis is presented to derive the moment generating functions of the probability distributions for packet departures of both voice and data traffic and for the data packet delay. For some cases with different numbers of voice codes, numerical results display the correlation coefficient of the voice and data packet departures and the coefficient of variation of the data packet delay as well as average performance measures, such as the throughput, the average delay of data packets, and the average blocking probability of voice calls.