Penalized regression with correlation-based penalty

A new regularization method for regression models is proposed. The criterion to be minimized contains a penalty term which explicitly links strength of penalization to the correlation between predictors. Like the elastic net, the method encourages a grouping effect where strongly correlated predictors tend to be in or out of the model together. A boosted version of the penalized estimator, which is based on a new boosting method, allows to select variables. Real world data and simulations show that the method compares well to competing regularization techniques. In settings where the number of predictors is smaller than the number of observations it frequently performs better than competitors, in high dimensional settings prediction measures favor the elastic net while accuracy of estimation and stability of variable selection favors the newly proposed method.     

A model selection criterion based on the BHHJ measure of divergence

The aim of this work is the discussion and investigation of measures of divergence and model selection criteria. A recently introduced measure of divergence, the so-called BHHJ measure (Basu, A., Harris, I.R., Hjort, N.L., Jones, M.C., 1998. Robust and efficient estimation by minimising a density power divergence. Biometrika 85, 549–559) is investigated and a new model selection criterion the divergence information criterion (DIC) based on this measure is proposed. Simulations are performed to check the appropriateness of the proposed criterion.     

鲜卑语言解读述论

本文用传统的考据学方法对白鸟库吉于 80年前解读的一批鲜卑词语进行了重新审订 ,并在其基础上补充了前人没有注意到的“若干”、“去斤”、“普”、“阿六敦”、“贺六浑”五个词 ,同时提出了利用鲜卑人名字来考释鲜卑词义的可能性     

明清时期徽州女子的择偶标准

明清时期,受到朱子《家礼》的影响,徽州宗族严格规定女子择偶需重视对方的家世门第与人品才学而贫富不论,形成了一套以门第为主、兼顾才学、不重财富的择偶标准,徽州女子择偶往往无自主权,全凭宗族家长决定。但透过森严的族规家法和官方文献,徽州女子也有着自己的一套择偶标准,家世才学、财富、外貌即是其中的重要内容。     

Factors Affecting Alcohol Purchase Decisions and Expenditures: A Sample Selection Analysis by Ethnicity in Malaysia

Heckman’s sample selection model was applied to data from the Malaysian Household Expenditure Survey 2004/2005 to examine the factors influencing the likelihood of purchasing and the amount spent on alcohol in Malaysia. Results of the marginal effects suggest that while socio-demographic factors are important determinants of household purchase decisions and expenditure levels on alcohol in Malaysia, the effects vary across ethnic groups. Specifically, although education had a significant but modest impact in reducing the probability of alcohol purchases and expenditure levels among ethnic Chinese households, this effect was not evident among the ethnic Indians and other races. While increasing household size lowered the likelihood of purchasing alcohol and its expenditure levels for all ethnic groups, the reinforcing effects of both income and gender were relevant only for ethnic Chinese and Indian households. Last, urban Indian households were less likely to purchase alcohol and spend less compared to rural Indian households.

Gender Asymmetry in Educational and Income Assortative Marriage

The reversal of the gender gap in education has reshaped the U.S. marriage market. Drawing on data from the 1980 U.S. Census and the 2008–2012 American Community Surveys, the author used log‐linear models to examine gender asymmetry in educational and income assortative mating among newlyweds. Between 1980 and 2008–2012, educational assortative mating reversed from a tendency for women to marry up to a tendency for women to marry down in education, whereas the tendency for women to marry men with higher incomes than themselves persisted. Moreover, in both time periods, the tendency for women to marry up in income was generally greater among couples in which the wife's education level equaled or surpassed that of the husband than among couples in which the wife was less educated than the husband. The author discusses the implications of the rising female advantage in education for gender change in heterosexual marriages.     

The Effect of External Influences on Mate Selection Necessity Traits: Cross-Cultural Comparisons of Chinese and American Men and Women

In the current study we conducted hierarchical regression analyses to examine the extent to which influences from parents, peers, and the media explained participants’ perceptions of their possible mates’ criteria and their own criteria for age, attractiveness, social status, and wealth in choosing a spouse. Participants included 333 Chinese individuals and 339 Americans. We found that the three external influences accounted for a significant amount of the variation in the perceptions of and the criteria for the four mate selection necessity traits. In particular, media influence was the most significant and consistent external influence predictor. We discussed the cultural and gender differences observed in this study and how our findings might fit into the existing mate selection literature.     

A Good Match? Offline Matchmaking Services and Implications for Gender Relations

Faced with barriers to successful coupling, namely disappointments with online dating, rising numbers of North Americans of varying ages and backgrounds are using personalized, offline matchmaking services to find long-term partners. However, few studies have examined the process interpretively from clients’ and matchmakers’ perspectives. Using interview data from 20 matchmakers and 10 heterosexual clients, content analyses of 102 company websites, and associated client comments and media coverage, this study queries connections between matchmaking’s growing popularity, (un)changing institutions, and gender relations. Analyses demonstrate that opportunities and constraints offered by the strategy are gendered, with men largely maintaining the partnering privileges they enjoy in other dating arenas and women making modest gains when participating as paying clients. Experiences are further shaped by age, ethnicity, and socioeconomic status.     

The characterization of Monte Carlo errors for the quantification of the value of forensic evidence

Recent developments in forensic science have lead to a proliferation of methods for quantifying the probative value of evidence by constructing a Bayes Factor that allows a decision-maker to select between the prosecution and defense models. Unfortunately, the analytical form of a Bayes Factor is often computationally intractable. A typical approach in statistics uses Monte Carlo integration to numerically approximate the marginal likelihoods composing the Bayes Factor. This article focuses on developing a generally applicable method for characterizing the numerical error associated with Monte Carlo integration techniques used in constructing the Bayes Factor. The derivation of an asymptotic Monte Carlo standard error (MCSE) for the Bayes Factor will be presented and its applicability to quantifying the value of evidence will be explored using a simulation-based example involving a benchmark data set. The simulation will also explore the effect of prior choice on the Bayes Factor approximations and corresponding MCSEs.