Outlier detection in high-dimensional regression model

An outlier is defined as an observation that is significantly different from the others in its dataset. In high-dimensional regression analysis, datasets often contain a portion of outliers. It is important to identify and eliminate the outliers for fitting a model to a dataset. In this paper, a novel outlier detection method is proposed for high-dimensional regression problems. The leave-one-out idea is utilized to construct a novel outlier detection measure based on distance correlation, and then an outlier detection procedure is proposed. The proposed method enjoys several advantages. First, the outlier detection measure can be simply calculated, and the detection procedure works efficiently even for high-dimensional regression data. Moreover, it can deal with a general regression, which does not require specification of a linear regression model. Finally, simulation studies show that the proposed method behaves well for detecting outliers in high-dimensional regression model and performs better than some other competing methods.     

Do individuals successfully cover up their lies? Evidence from a compliance experiment

We study how well individuals in a face-to-face situation can delude others. We exploit data from a laboratory experiment in which participants were asked to assess video-taped statements as being rather truthful or untruthful. The statements are face-to-face tax declarations. The video clips feature each subject twice making the same declaration: One time the subject is reporting honestly, and the other time willingly dishonestly. This allows us to investigate within-subject differences in perceived truthfulness. Our study provides several novel insights. We find that individuals can cover up their lies successfully. On average, a subject is perceived as more truthful if she deceives than if she reports honestly. In particular, individuals with a genuine dishonest appearance manage to increase their perceived truthfulness by up to 14 percent when lying. Moreover, our results show that a subject appears less truthful if she had previously been caught lying. Being detected as a liar previously appears to impair self-confidence and to thereby lower an individual’s ability to deceive.     

Investment communities: Behavioral attitudes and economic dynamics

Using a real-world data set encompassing the daily portfolio holdings and exposures of complex investment funds, we derive a set of quantitative attributes to capture essential behavioral features of fund managers. We find the existence and stability of three investment attitudes, namely the conservative, the reactive, and the pro-active profiles, defining communities that respond differently when facing external shocks. The conservative community has behavioral similarities that tend to decrease due to external shocks, the reactive community members greatly increase their activity level especially during turmoil phases, while delegated investors in the pro-active community are more resilient to turbulence and counterbalance the impact of the events by adjusting their portfolio exposures in advance. We show that exogenous shocks only temporarily perturb the behavioral traits of the communities which then go back to their original states once the distress is embedded.     

Resource Distribution in Multiple Attacks with Imperfect Detection of the Attack Outcome

This article extends the previous research of consecutive attacks strategy by assuming that an attacker observes the outcome of each attack imperfectly. With given probabilities it may wrongly identify a destroyed target as undestroyed, and wrongly identify an undestroyed target as destroyed. The outcome of each attack is determined by a contest success function that depends on the amount of resources allocated by the defender and the attacker to each attack. The article suggests a probabilistic model of the multiple attacks and analyzes how the target destruction probability and the attacker's relative resource expenditure are impacted by the two probabilities of incorrect observation, the attacker's and defender's resource ratio, the contest intensity, the number of attacks, and the resource distribution across attacks. We analyze how the attacker chooses the number of attacks, the attack stopping rule, and the optimal resource distribution across attacks to maximize its utility.     

Mean shift and influence measures in linear measurement error models with stochastic linear restrictions

We present influence diagnostics for linear measurement error models with stochastic linear restrictions using the corrected likelihood of Nakamura in 1990. The case deletion and mean shift outlier models are developed to identify outlying and influential observations. We derive a corrected score test statistic for outlier detection based on mean shift outlier models. The analogs of Cook's distance and likelihood distance are proposed to determine influential observations based on case deletion models. A parametric bootstrap procedure is used to obtain empirical distributions of the test statistics and a simulation study has been used to evaluate the performance of the proposed estimators based on the mean squares error criterion and the score test statistic. Finally, a numerical example is given to illustrate the theoretical results.     

Adolescents’ perceptions of the ‘substance use-violence nexus’: a South African perspective

The aim of the study was to explore adolescents’ perceptions of substance use as a contributing factor to community violence by employing a conceptual framework proposed by Goldstein. Data were collected from a sample of adolescents between the ages of 15 and 16 years by means of focus group interviews. Theoretical thematic analysis was used to analyse and interpret the findings. The participants’ perceptions indicated that substance use and withdrawal symptoms induce physiological and behavioural modifications resulting in aggression and violence. The findings also suggest that participants believed that substance users frequently commit property and violent interpersonal crimes such as theft, robbery, assault and murder to procure their substances. Further, the adolescents’ perceived violence to be rooted within gang culture as well as the involvement in shared markets for illicit substance trading.     

Bid-rigging networks and state-corporate crime in the construction industry

This study highlights the symbiotic relationship that may arise between political and business actors in conspiracies organized around public construction bids. Based on a core-periphery social network analysis, the study tracks the emergence of such a conspiracy in the city of Laval (Canada) by illustrating the evolution of bid-rigging networks revolving around suspected and persistent acts of corruption, bid-rigging, and bribery. To assess this process, the study monitors irregular bidding indicators across a data set compiled from more than 7000 public construction tenders that were processed by the city from 1966 to 2013. Findings reveal that firms suspected of bid-rigging activities were perennial core participants largely as a result of a state-corporate crime system that served as the guiding force for agreements between the main construction entrepreneurs.     

Composite Estimation for Single‐Index Models with Responses Subject to Detection Limits

We propose a semiparametric estimator for single‐index models with censored responses due to detection limits. In the presence of left censoring, the mean function cannot be identified without any parametric distributional assumptions, but the quantile function is still identifiable at upper quantile levels. To avoid parametric distributional assumption, we propose to fit censored quantile regression and combine information across quantile levels to estimate the unknown smooth link function and the index parameter. Under some regularity conditions, we show that the estimated link function achieves the non‐parametric optimal convergence rate, and the estimated index parameter is asymptotically normal. The simulation study shows that the proposed estimator is competitive with the omniscient least squares estimator based on the latent uncensored responses for data with normal errors but much more efficient for heavy‐tailed data under light and moderate censoring. The practical value of the proposed method is demonstrated through the analysis of a human immunodeficiency virus antibody data set.     

律师伪证罪的困境与出路

《刑法》第306条有关律师伪证罪的规定一直面临着悖逆民心的困境，但其原因并非实体内容不公正，而是欠缺程序上的合理限制，由此导致它沦为检察机关实施职业报复的工具。将律师纪律处分或行政处罚程序作为刑事诉讼程序的程序性前提，把取消会员资格的行业处分或吊销律师职业证书的行政处罚作为刑事诉讼的实体性前提，并限制刑事侦查的调查范围，可以合理地限制检察机关的偏见和恣意，带领该规定走出当前的困境。