Technical Sessions – Monday July 11
Technical Sessions – Monday July 11
Technical Sessions – Monday July 11
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
MB-12 IFORS 20<strong>11</strong> - Melbourne<br />
4 - Light-Tailed Asymptotics of the Joint Queue Length<br />
Distribution of a Multiclass M/G/1 Queue with the FCFS<br />
Service Discipline<br />
Kenzo Hara, Dept. of Systems Science, Graduate School of<br />
Informatics, Kyoto University, 606-8501, Kyoto, Japan,<br />
ken3@sys.i.kyoto-u.ac.jp, Hiroyuki Masuyama, Yutaka<br />
Takahashi<br />
This talk is concerned with a multiclass M/G/1 queue with the FCFS service<br />
discipline, which is fed by independent Poisson arrival processes with different<br />
service time distributions. For the multiclass M/G/1 queue, it is, in general,<br />
difficult to obtain an explicit expression of the stationary joint queue length distribution.<br />
We derive a light-tailed asymptotic formula for the stationary joint<br />
queue length distribution, assuming that the service time distributions are all<br />
light-tailed. The asymptotic formula is useful to understand the tail behavior<br />
and expected to serve as an approximation.<br />
� MB-12<br />
<strong>Monday</strong>, 14:00-15:30<br />
Meeting Room 205<br />
Services Management<br />
Stream: Contributed Talks<br />
Contributed session<br />
Chair: Ronsen Purba, Information Technology, STMIK Mikroskil,<br />
Jalan Thamrin No.124/140 Medan, Jalan Sembada XI No. 9, 20212,<br />
Medan, North Sumatera, Indonesia, ronsen@mikroskil.ac.id<br />
1 - A Case Study of Design and Implement of a Service<br />
Business on a Heavy Machinery Manufacturing Firm<br />
Nobuhiko Nishimura, Faculty of Economics, Nagasaki<br />
University, 4-2-1, Katafuchi, 850-8506, Nagasaki, Japan,<br />
nishimra@nagasaki-u.ac.jp<br />
Three major perspectives relating to service business are reviewed, followed by<br />
an in-depth case study on a new e-service business design and implementation<br />
initiative at a heavy machinery manufacturing firm. It is confirmed that profitable<br />
services business can be designed based on strategic fitting to capability<br />
and position already built by the firm in its conventional business, with repeat<br />
prototyping to elicit potential customers’ requirements. The resulting system<br />
makes the service offering complementary to and dependent upon products to<br />
raise the entry barrier against competitors.<br />
2 - Verification of e-customer Purchasing Behavior Pattern<br />
Model by Cohort Analysis<br />
Hisaya Sunada, Web Business Technology, The Kyoto College of<br />
Graduate Studies for Informatics, Japan,<br />
hisaya.sunada@gmail.com, Tong Wu, Wang Chao, Maotao<br />
Chen, Hongxue Wang, Hong Seung Ko<br />
In the e-business environment, the visualization of e-customer purchasing behavior<br />
is absolutely necessary to select the most valuable e-customer who must<br />
be retained. In this paper, we perform the cohort analysis to the segmentation<br />
factor, that is purchasing frequency ,profitable sales and 7 Steps of e-customer<br />
purchasing behavior process in the e-customer purchasing behavior pattern<br />
model which is proposed by Ko et al. Consequently, we check up the validity<br />
of the e-customer purchasing behavior pattern model by the characteristics<br />
on the e-customer purchasing behavior pattern as a result.<br />
3 - A use of Bootstrapped Malmquist Indices to Assess<br />
Productivity Changes of the Iranian Banks<br />
16<br />
Amir Arjomandi, Economics, University of Wollongong, 5/20,<br />
Smith Street, 2522, Wollongong, NSW, Australia,<br />
amira@uow.edu.au, Abbas Valadkhani<br />
This study employs various bootstrapped Malmquist indices and efficiency<br />
scores to investigate the effects of government regulation on the performance of<br />
the Iranian banking industry over the period 2003-2008. An alternative decomposition<br />
of the Malmquist index, introduced by Simar and Wilson (1998a), is<br />
also applied to decompose technical changes further into pure technical change<br />
and changes in scale efficiency. A combination of these approaches facilitates<br />
a robust and comprehensive analysis of Iranian banking industry performance.<br />
While this approach is more appropriate than the traditional Malmquist approach,<br />
for the case of banking efficiency studies, it has not previously been<br />
conducted for any developing country’s banking system. The results obtained<br />
show that although, in general, the regulatory changes had different effects on<br />
individual banks, the efficiency and productivity of the overall industry declined<br />
after regulation. We also find that productivity had positive growth before<br />
regulation mainly due to improvements in pure technology, and that government<br />
ownership had an adverse impact on the efficiency level of state-owned<br />
banks. The bootstrap approach demonstrates that the majority of estimates obtained<br />
in this study are statistically significant.<br />
4 - On Deciding Sustainable Distribution Centre of a Supply<br />
Chain<br />
Ronsen Purba, Mathematics Dept., STMIK Mikroskil<br />
Medan/Graduate School of Mathematics, University of Sumatera<br />
Utara, Jalan Thamrin No. 140 Medan North Sumatera, Medan,<br />
Indonesia, purbaronsen17@yahoo.com<br />
A supply chain is a network that performs the procurement of raw material, the<br />
transportation of raw material to intermediate and end products, and the distribution<br />
of end products to retailers or customers. In this paper we address a new<br />
approach for solving a supply chain optimization model to decide distribution<br />
centres based on environmental consideration.<br />
� MB-13<br />
<strong>Monday</strong>, 14:00-15:30<br />
Meeting Room 206<br />
Data Mining and Continuous Optimization<br />
Stream: Continuous and Non-Smooth Optimization<br />
Invited session<br />
Chair: Zeev (Vladimir) Volkovich, Ort Braude Academic College,<br />
Yehiam 6, 21823, Karmiel, Israel, zeev@actcom.co.il<br />
1 - Necessary Conditions for the Confidence Level of the<br />
Randomized Cluster Stability Algorithm<br />
Oleg Granichin, Mathematics & Mechanics, Sankt-Petersburg<br />
State University, 28 Universitetskii prospect, Petergof, 198504,<br />
St. Petersburg, Russian Federation, Oleg_granichin@mail.ru,<br />
Renata Avros, Zeev (Vladimir) Volkovich, Mikhail Morozkov<br />
The well known machine learning problem is an estimation of clusters amount<br />
in a given dataset. We offer the approach in the framework of the common<br />
"elbow" methodology such that the true number of clusters is recognized as the<br />
discontinuity point of the differential risk function. We apply a randomized algorithm<br />
to allocate this position. The scenario approach is used to significantly<br />
reduction the computation complexity. We present necessary conditions to provide<br />
the chosen by user level of confidence. Numerical simulation example of<br />
an unknown huge amount of clusters illustrates the theory.<br />
2 - Cluster Analysis of A Tobacco Control Data Set<br />
Zari Dzalilov, School of Information Technology and<br />
Mathematical Sciences, University of Ballarat, 1, University<br />
Drive, 3353, Ballarat, VIC, Australia,<br />
z.dzalilov@ballarat.edu.au, Adil Bagirov<br />
Development of theoretical and methodological frameworks in data analysis<br />
is fundamental for modeling complex tobacco control systems. Optimization<br />
methods can detect nonlinearity and be effective analysis tools of complex data.<br />
We evaluate the modified global k-means clustering algorithm for a massive set.<br />
Cluster analysis identified fixed and stable clusters in the studied data. Clusters<br />
correspond to smoker groups with similar behavior, identification of these clusters<br />
may allow modification of existing tobacco control systems and on design<br />
of future data acquisition surveys.<br />
3 - Algorithm Learning Based Neural Network for Multivariate<br />
Classification<br />
Hyunsoo Yoon, School of Industrial Management Engineering,<br />
Korea University, Anam 5-ga, Anam-dong, Sungbuk-gu,