Kernel Modification Effects for Support Vector Machine Applied to Limit Order Book of Nikkei 225 Futures
Hayato Kijima1, Hideyuki Takada2
1. Graduate School of Information Science, Toho University, Miyama 2-2-1, Chiba 274-8510, Funabashi, Japan.
2. Department of Information Science, Toho University,Miyama 2-2-1, Chiba 274-8510, Funabashi, Japan.
Abstract: Market participants place their limit/market orders by taking into account both the trajectory and current status of the limit order book. This behavior is based on the policy that the shape of the limit order book is quite informative for predicting future direction of a traded asset. In this paper, we employ Support Vector Machine to learn future mid-price directions and apply conformal transformation of the kernel function in order to improve its accuracy. Our empirical studies are based on Nikkei 225 futures and show that the conformal transform methods improved the precision more than 3% in average compared to the standard Gaussian RBF kernel. We further investigate numerically how the precision is improved by controlling parameter involved in the conformal transform.
Keywords: Limit Order Book, Support Vector Machine, Riemannian metric,Conformal transformation.
Pages: 149 – 167 | Full PDF Paper
Comparison of the Performance of Inpatient Care for Chemotherapy Patients in RSUP Dr. Hasan Sadikin Bandung West Java Using Queuing Theory
MuthiaNadhira Faladiba1,a), NikenNastia N.N.2,b), and Atina Ahdika3,c)
1, 2. Students of Department of Statistics, Islamic University of Indonesia, JalanKaliurang Km 14.5 Sleman Yogyakarta 55584.
3. Lecturer at Department of Statistics, Islamic University of Indonesia.
Abstract: Waiting is the underlying presence of a queue. The queue process is a process associated with the arrival of customers at a care facility, waiting in the queue if the line cannot be served, being served, and eventually left the facility after being served. This article studied the queue models and customer care processes of inpatient chemotherapy in RSUP Dr. HasanSadikin Bandung West Java. The analysis was performed by determining the probability distribution of the arrival and service time using One Sample Kolmogorov-Smirnov test, specify a model queue for each class, and determine the effectiveness of patient care through the calculation of performance measures of queuing model in the class of hospitalization. The queue models which obtained from the analysis are M/G/S model for inpatient class 1, 3, and VIP, and M/M/S model for inpatient class 2. The result shows that the system has been effective in each class based on the values of each performance measurements.
Keywords: chemotherapy, distribution test, inpatient, performance measurements, queue models.
Pages: 168 – 178 | Full PDF Paper
Frailty Modeling for Repairable Systems with Minimum Repair: An Application to Dump Truck Data of a Brazilian Mining Company
Amanda Morales Eudes D’Andrea1, Cirdêmia Costa Feitosa1, Vera Lucia Damasceno Tomazella2, Afrânio Márcio Corrêa Vieira2
1. Universidade Federal de São Carlos, Departamento de Estatística, São Carlos, SP, Brasil e Universidade de São Paulo, Instituto de Ciências Matemáticas e de Computação, São Carlos, SP, Brasil.
2. Universidade Federal de São Carlos, Departamento de Estatística, São Carlos, SP, Brasil.
Abstract: In repairable system data analysis it is common to many components of the same type are studied and in these cases it is relevant to verify the heterogeneity between systems. Proschan  pointed out that the unobserved heterogeneity may explain increasing failure rates, which is often found in reliability analysis. The unobserved heterogeneity may be estimated from models of frailty. This model is characterized by using a random effect, that is, a non-observable random variable that represents the information that could not or were not observed. However, according with Vaupel et al. , the standard methods in repairable system data analysis ignore the unobserved heterogeneity. Thus, this work will explore the frailty models. The inferential method for estimation of the parameters will be displayed for models with minimal repair. Finally, an application to real data set  was taken, in which the models with and without frailty are compared. These real data set involving failures in trucks was collected in a Brazilian mining company.
Keywords: Repairable systems, minimal repair, Power Law Process, frailty.
Pages: 179 – 198 | Full PDF Paper