論文

This study addresses a fundamental, yet overlooked, gap between standard theory and empirical modelling practices in the OLS regression model y=Xβ+u with collinearity. In fact, while an estimated model in practice is desired to have stability and efficiency in its “individual OLS estimates”, y itself has no capacity to identify and control the collinearity in X and hence no theory including model selection process (MSP) would fill this gap unless X is controlled in view of sampling theory. In this paper, first introducing a new concept of “empirically effective modelling” (EEM), we propose our EEM methodology (EEM-M) as an integrated process of two MSPs with data (yo,X) given. The first MSP uses X only, called the XMSP, and pre-selects a class D of models with individually inefficiency-controlled and collinearity-controlled OLS estimates, where the corresponding two controlling variables are chosen from predictive standard error of each estimate. Next, defining an inefficiency-collinearity risk index for each model, a partial ordering is introduced onto the set of models to compare without using yo, where the better-ness and admissibility of models are discussed. The second MSP is a commonly used MSP that uses (yo,X), and evaluates total model performance as a whole by such AIC, BIC, etc. to select an optimal model from D. Third, to materialize the XMSP, two algorithms are proposed with applications.

We propose a novel framework to investigate lead-lag relationships between two financial assets. Our framework bridges a gap between continuous-time modeling based on Brownian motion and the existing wavelet methods for lead-lag analysis based on discrete-time models and enables us to analyze the multiscale structure of lead-lag effects. We also present a statistical methodology for the scale-by-scale analysis of lead-lag effects in the proposed framework and develop an asymptotic theory applicable to a situation including stochastic volatilities and irregular sampling. Finally, we report several numerical experiments to demonstrate how our framework works in practice.

As of the end of October 2017, the utilization of ETC 2.0 in Japan is about 15%, which is growing at a faster rate than its predecessor ETC1.0. ETC 2.0 probe data is designed to store new data fields, or features, such as position information history of the vehicle and abrupt driving maneuver history. Up to the present, the use of these data has not been well spreading, and the government of Japan supports the utilization of ETC 2.0 by the administrative and private enterprises. In this study, we aim to explore a new, effective use of ETC 2.0 probe data. The probe data is sizable, accumulating about 400 GB for one month. Analysis is conducted focusing on the number of abrupt driving maneuver in ten most popular tourist areas in the Greater Tokyo Area ('Kanto region') from April 2015 to March 2016. Using the ETC 2.0 probe data, we propose a methodology to quantify abrupt driving maneuver of individual drivers and to evaluate their driving behaviors with a view towards business applications in mind.

This study is concerned with the'Internet meme', an image with a brief message attached to a Social Network Service (SNS) post expressing one's mood humorously. Internet memes have become popular especially among young users recently. Meme marketing, one of the growing corporate strategies designed for more effective online brand communication, has been evolving in environments where younger generations have become the most active individuals and constitute the majority of the virtual Internet society of today. The purpose of this study is to quantify the effects of Internet memes in the context of social media marketing on SNS platforms. A/B testing is conducted with dozens of post experiments on an SNS platform that has the most users in China to test the effect of Internet memes by measuring the post engagement rates. Further, a statistical analysis using generalized linear mixed models is conducted for exploring the characteristics of users who tend to react to Internet memes. Practical implications for SNS marketing are also drawn.