scholarly journals Nonparametric regression with adaptive truncation via a convex hierarchical penalty

Biometrika ◽  
2018 ◽  
Vol 106 (1) ◽  
pp. 87-107
Author(s):  
Asad Haris ◽  
Ali Shojaie ◽  
Noah Simon
Keyword(s):  
Nonparametric Regression ◽  
Empirical Studies ◽  
Finite Basis ◽  
Additive Models ◽  
Computationally Efficient ◽  
Penalized Estimation ◽  
Minimax Rates ◽  
Minimax Rate ◽  
Finite Basis Representation ◽  
Different Levels

SUMMARY We consider the problem of nonparametric regression with a potentially large number of covariates. We propose a convex, penalized estimation framework that is particularly well suited to high-dimensional sparse additive models and combines the appealing features of finite basis representation and smoothing penalties. In the case of additive models, a finite basis representation provides a parsimonious representation for fitted functions but is not adaptive when component functions possess different levels of complexity. In contrast, a smoothing spline-type penalty on the component functions is adaptive but does not provide a parsimonious representation. Our proposal simultaneously achieves parsimony and adaptivity in a computationally efficient way. We demonstrate these properties through empirical studies and show that our estimator converges at the minimax rate for functions within a hierarchical class. We further establish minimax rates for a large class of sparse additive models. We also develop an efficient algorithm that scales similarly to the lasso with the number of covariates and sample size.

scholarly journals Beyond a Dichotomous Understanding of Online Anonymity: Bridging the Macro and Micro Level

2021 ◽  
pp. 136078042110197
Author(s):  
Lina Eklund ◽  
Emma von Essen ◽  
Fatima Jonsson ◽  
Magnus Johansson
Keyword(s):  
Empirical Studies ◽  
Online Auctions ◽  
Future Research ◽  
Plural Form ◽  
Contentious Issue ◽  
Online Anonymity ◽  
Complex Process ◽  
Theoretical Synthesis ◽  
Micro Level ◽  
Different Levels

Anonymity on the Internet is a contentious issue; by some seen as an important freedom to be protected, while others argue for increased identification to protect groups at risk of exploitation. The debate reflects a dichotomous view of online anonymity; you are, or you are not anonymous. However, anonymity is a complex process played out on different levels and defined by various actors. While empirical studies show this, theoretical synthesis is lacking. This essay provides perspective on anonymity online by comparing two critical cases, online auctions and online gaming, we corroborate results from a 4-year interdisciplinary project with researchers from sociology, economics, and computer and system sciences. We argue that one should talk about anonymities in plural form, as online anonymity is not a state but a relational process. We put forth a conceptual model, which unpacks online anonymity as interdependent macro structures – legal, commercial, and technological – and micro/meso facets – factual, social group, and physical – to be used in future research.

scholarly journals Optimal minimax rates against non-smooth alternatives

2021 ◽  
Author(s):  
Kohtaro Hitomi ◽  
Masamune Iwasawa ◽  
Yoshihiko Nishiyama
Keyword(s):  
Alternative Hypothesis ◽  
Error Variance ◽  
Smooth Functions ◽  
Engel Curves ◽  
Minimax Rates ◽  
Instrumental Variable Regression ◽  
Bounded Functions ◽  
The Difference ◽  
Set Up ◽  
Minimax Rate

Abstract This study investigates optimal minimax rates for specification testing when the alternative hypothesis is built on a set of non-smooth functions. The set consists of bounded functions that are not necessarily differentiable with no smoothness constraints imposed on their derivatives. In the instrumental variable regression set up with an unknown error variance structure, we find that the optimal minimax rate is n−1/4, where n is the sample size. The rate is achieved by a simple test based on the difference between non-parametric and parametric variance estimators. Simulation studies illustrate that the test has reasonable power against various non-smooth alternatives. The empirical application to Engel curves specification emphasizes the good applicability of the test.

scholarly journals Construction, Detection, and Interpretation of Crime Patterns over Space and Time

2020 ◽  
Vol 9 (6) ◽  
pp. 339
Author(s):  
Zengli Wang ◽  
Hong Zhang
Keyword(s):  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Empirical Studies ◽  
Prevention Strategies ◽  
Crime Patterns ◽  
Risk Levels ◽  
Research Areas ◽  
Research Objects ◽  
Crime Risk ◽  
Different Levels

Empirical studies have focused on investigating the interactive relationships between crime pairs. However, many other types of crime patterns have not been extensively investigated. In this paper, we introduce three basic crime patterns in four combinations. Based on graph theory, the subgraphs for each pattern were constructed and analyzed using criminology theories. A Monte Carlo simulation was conducted to examine the significance of these patterns. Crime patterns were statistically significant and generated different levels of crime risk. Compared to the classical patterns, combined patterns create much higher risk levels. Among these patterns, “co-occurrence, repeat, and shift” generated the highest level of crime risk, while “repeat” generated much lower levels of crime risk. “Co-occurrence and shift” and “repeat and shift” showed undulated risk levels, while others showed a continuous decrease. These results outline the importance of proposed crime patterns and call for differentiated crime prevention strategies. This method can be extended to other research areas that use point events as research objects.

Consistency at different levels

2020 ◽  
Vol 11 (4) ◽  
pp. 537-567 ◽  
Author(s):  
Tao Scofield Su ◽  
Chunhua Chen ◽  
Xiaoyu Cui ◽  
Chunsheng Yang ◽  
Weimo Ma
Keyword(s):  
Organizational Performance ◽  
Empirical Studies ◽  
Task Type ◽  
Measurement Tool ◽  
Regional Culture ◽  
Content Type ◽  
Positive Correlation ◽  
Vertical Distance ◽  
And Performance ◽  
Different Levels

Purpose This paper aims to answer following three important but not well-answered or unanswered questions in the extant trust literatures: What is the true magnitude that trust impacts on performance? Is there any consistency among the effects of trust on performance at different levels? How does vertical distance affect the trust-performance relationship? Design/methodology/approach It captures the law between trust and performance at different levels by conducting a meta-analytic examination consisting of 238 independent empirical studies, 586 effect sizes and 110,576 independent samples. Findings It makes a periodic conclusion that trust significantly promotes performance. Specifically, trust not only has stronger positive correlation with team performance than individual and organizational performance inside organization, but also strongly facilitates organizational performance between organizations. Moreover, consistency exits in the effects of trust on performance at different levels. On one hand, trust has stronger positive correlation with performance of contextual type than performance of innovative type than performance of task type at different levels. On the other hand, promotion effect of trust on performance strengthens when the vertical distance between trustors and trustees diminishes. Additionally, three potential moderators including publication status, measurement tool and common method variance moderate the focused relation, but moderating effect is not thorough for regional culture. Moderating directions of the above four potential moderators are highly consistent. Originality/value This paper answers the three important but not well-answered or unanswered questions.

Design and Evaluation of Extensions to UML Sequence Diagrams for Modeling Multithreaded Interactions

2009 ◽  
Vol 8 (2) ◽  
pp. 120-136 ◽  
Author(s):  
Shaohua Xie ◽  
Eileen Kraemer ◽  
R. E. K. Stirewalt ◽  
Laura K Dillon ◽  
Scott D Fleming
Keyword(s):  
Observational Study ◽  
Comparative Study ◽  
Teaching And Learning ◽  
Empirical Studies ◽  
Sequence Diagram ◽  
Sequence Diagrams ◽  
Multiple Thread ◽  
Uml Sequence Diagram ◽  
Different Levels

Learning about concurrency and synchronization is difficult for novices. Our research seeks to support and improve the teaching and learning of concurrency concepts and to improve comprehension of the intricacies of multiple thread interactions. This paper describes a series of empirical studies in the first phase of our research. We began by conducting a comparative study to empirically evaluate the usability by novices of the existing variants of the UML sequence diagram notation in solving comprehension tasks involving multiple thread interactions. The results implied that a deliberately designed variant of this notation may provide better support for reasoning about concurrent behavior. We then investigated the factors that complicate learning, with the idea that the same complexities would also complicate comprehension tasks. In order to understand the practical difficulties novices encounter in learning about concurrency, we conducted an instructor interview and an observational study. These investigations guided us in determining the desirable properties of a new notation. We then designed synchronization-adorned UML (saUML) sequence diagrams, which extend UML sequence diagrams with those properties. Finally, we performed four empirical studies to evaluate the usability and efficacy of saUML. Through these empirical studies, we were able to validate the benefits of saUML in enhancing novices' understanding of programs with different levels of synchronization complexity.

scholarly journals The Effectiveness of Different Levels of Activation in Higher Education

2018 ◽  
Author(s):  
Paula Figas ◽  
Michael Sailer ◽  
Georg Hagel ◽  
Maximilian Sailer
Keyword(s):  
Higher Education ◽  
Empirical Studies ◽  
Learning Performance ◽  
Test Results ◽  
Classroom Activities ◽  
Open Research ◽  
Significant Difference ◽  
Research Questions ◽  
Positive Effect ◽  
Different Levels

Although there are several empirical studies about active learning in higher education, there are some open research questions. Especially, it appears to be relevant to find out what type of activity (active, interactive and constructive) has a positive effect on the learning result of participants in task-based activities. In order to answer this question an experimental study was conducted with students in German higher education (N=50). The results do not show a significant difference between active, interactive and constructive activities on learning performance in task-based learning. But all groups have benefited from the classroom activities according to the test results.

Sign in / Sign up

Export Citation Format

Share Document