Manyo tanka is a short Japanese poem included in Manyoshu which is the oldest collection of Japanese poems. Since the short poems are composed by several poets, each poem has characteristics for each poet. Until now, the characteristics have been subjectively studied or have been investigated based on a single sound. In this paper, we use a statistical method to study the characteristics based on multiple sounds in Manyo tanka. In particular, we analyze the Manyo tanka dataset using sparse canonical discriminant analysis. This analysis uncovers inherent properties of poets for Manyo tanka.

Production of components in ad creative itself is largely dependent on experience of designer. When you make creatives, it is possible to improve production process more efficiently if there is criteria or standard that guides you what components are more important. In this study, we conducted an empirical analysis to measure contribution of components in ad creative in the framework of CTR prediction for mobile advertising. By computer vision technology, human-interpretable keywords and color information were extracted as components that configure ad creative. As learner for CTR prediction, we identified which components in ad creative are effective for click by estimating importance and interaction of each feature value from the result of GBDT (Gradient Boosted Decision Trees) using a decision tree as weak classifier. The estimation of interaction was based on Interpretable Trees (inTrees) by Deng (2019). By combining computer vision technology and machine learning methods that can estimate importance and interaction of feature value, it enables not only to measure components of ad creative and their contributions, but also to expect wide range of applications.

Quality can be divided into objective elements, such as material characteristics, and subjective elements, such as beauty. The latter is called perceived quality, and in recent years, it has been emphasized more than the objective elements. In the automobile industry, perceived quality is one of the most important determinants of competitiveness. However, being a subjective concept, the estimation of its quantitative effect is insufficient. The purpose of this study is to evaluate the influence of the perceived quality of car exteriors on the customers' willingness to pay (WTP) by conducting a randomized controlled trial (RCT). In the RCT, participants were randomly assigned to two groups, with the homogeneity of each group confirmed by including items such as gender, age, household income, and so on in the recruiting survey. WTP was measured using the contingent valuation method to evaluate responses to an open-ended question. Among ``Color, Material, and Finish (CMF)''---the constituents of perceived quality---Finish was focused on. Evaluation through the Brunner-Munzel test showed a significant difference in the results. Although quantitative evaluation using statistical methods has mainly focused on objective elements of products, this study shows that it is possible to evaluate even the sensory elements of design finish from a consumer perspective, such as WTP.

The hybrid lognormal distribution (Kumazawa, Ohashi 1986) is a probability distribution generated by the existence of a mechanism that suppresses the frequency of occurrence of higher values predicted by lognormal. This is a distribution with a main body of lognormal and a right tail of normal distribution and was derived as a risk management model of a stochastic process that reasonably suppresses dose accumulation accompanying radiation work. Similar probability distributions are widely found in natural phenomena, engineering, economics, culture, sports and social statistics. In this paper, we premised that ``the rational adjustment of increase and suppression of quantity'' of the mechanism which generates the hybrid lognormal distribution is feasible to become a universal risk management formula and that the risk output from this management formula becomes a hybrid fluctuation (combining the logarithmic fluctuation and the linear fluctuation). Then we presented the concept and application examples in terms of identifying the hybrid fluctuation on a hybrid scale (an integration of logarithmic scale and linear scale) as well as on a two-dimensional graph paper having a hybrid scale on both axes. In the application example, a hybrid scale (HS) model that gives a bestlinear graph on hybrid-hybrid graph paper was applied to therelationship between protons and excess neutrons, where the strongCoulomb repulsion of protons in the stable isotopes of heavier nucleiis moderated by an excess of neutrons with strong attractiveinteractions. We have verified that the nuclear structural stabilitywith respect to neutron excess is the same as the risk managementformula derived from dose management.