All issues
Volume 44 (2015)
- Issue 3 Pages 71-
- Issue 1-2 Pages 1-
Volume 44, Issue 1-2
Displaying 1-4 of 4 articles from this issue
- |<
- <
- 1
- >
- >|
Preface
-
Tomoyuki HiguchiArticle type: Preface
2015 Volume 44 Issue 1-2 Pages 1-2
Published: 2015
Released on J-STAGE: January 22, 2016
JOURNAL OPEN ACCESSDownload PDF (514K)
Forums
-
Nobuhiko TeruiArticle type: Forums
2015 Volume 44 Issue 1-2 Pages 3-15
Published: 2015
Released on J-STAGE: January 22, 2016
JOURNAL OPEN ACCESSThis article surveys statistical analysis and modeling by defining marketing in several domains according to type and scale of marketing data. Then I discuss the role and perspective of statistics in marketing in the age of Big data.View full abstractDownload PDF (1364K) -
Hiroe TsubakiArticle type: Forums
2015 Volume 44 Issue 1-2 Pages 17-30
Published: 2015
Released on J-STAGE: January 22, 2016
JOURNAL OPEN ACCESSThe Forum will discuss the basic viewpoints required to model input-output systems or dominant rules for explaining the output of business in several input variables. For this purpose, illustrative financial data analyses are used as a case study. First, we will introduce that static law of business is described in the form that contains the error to a simple proportional relationship such as labor productivity. Then, the law should be noted that it is modified through appropriate classification. In addition, for description of the dynamic law, it shows that a simple approach is to model the structure of the historical difference data of inputs and the corresponding output. Though the residual distribution after fitting these models may be usually subject to fat tailed distribution, appropriate deletion of the outlying data group will be necessary and useful in case that variation of the outlying data that are considered as the data from frontiers of high or low business performances is caused by systematic effects of the original input variables. The rank logit model can be utilized to identify such systematic effects effectively. Finally, it is possible to extend a single equation model to simultaneous equation models to describe more realistic business models contributing to the improvement of the key performance indices including the financial performance as revenue.View full abstractDownload PDF (1345K)
Articles
-
Masaaki Tsujitani, Yusuke TanakaArticle type: Articles
2015 Volume 44 Issue 1-2 Pages 31-48
Published: 2015
Released on J-STAGE: January 22, 2016
JOURNAL OPEN ACCESSBone marrow transplants are a standard treatment for acute leukemia. Recovery following bone marrow transplantation is a multi-state process. This article examines predictions of probabilities at some points in multi-state survival models. Cox's proportional hazards model has been widely used for the analysis of treatment and prognostic effects with censored survival data. The model was developed based on the relation between survival and the patient characteristic observed when the patient entered the study. When the covariates values change for the duration of the study, however, some theoretical problems to be solved with respect to baseline survival function and baseline cumulative hazard function are involved. Several prognostic models have become widespread using the Cox's proportional hazards modes for the analysis of survival data having time-dependent covariates. In the present study, we propose nonlinear analysis based on generalized additive models with B-spline to predict the survival for the following short-term (say, 1 year) at any time during the course of the disease, and select the optimum smoothing parameters based on a valiant multifold cross-validation method. In order to summarize the measure of goodness-of-fit, the deviance on fitting a generalized additive model can be bootstrapped.View full abstractDownload PDF (1084K)
- |<
- <
- 1
- >
- >|