scholarly journals Erratum to: The impact of statistical learning on violations of the sure-thing principle

2015 ◽  
Vol 50 (2) ◽  
pp. 117-117
Author(s):  
Nicky Nicholls ◽  
Aylit Tina Romm ◽  
Alexander Zimper
Keyword(s):  
Statistical Learning ◽  
Sure Thing Principle ◽  
The Impact

scholarly journals The impact of statistical learning on violations of the sure-thing principle

2015 ◽  
Vol 50 (2) ◽  
pp. 97-115 ◽  
Author(s):  
Nicky Nicholls ◽  
Aylit Tina Romm ◽  
Alexander Zimper
Keyword(s):  
Statistical Learning ◽  
Sure Thing Principle ◽  
The Impact

scholarly journals Statistical Learning with a Nuisance Component (Extended Abstract)

2020 ◽  
Author(s):  
Dylan J. Foster ◽  
Vasilis Syrgkanis
Keyword(s):  
Statistical Learning ◽  
Domain Adaptation ◽  
Estimation Error ◽  
Nuisance Parameter ◽  
Individual Performance ◽  
Excess Risk ◽  
Risk Minimization ◽  
Target Parameter ◽  
Estimation Algorithms ◽  
The Impact

We provide excess risk guarantees for statistical learning in a setting where the population risk with respect to which we evaluate a target parameter depends on an unknown parameter that must be estimated from data (a "nuisance parameter"). We analyze a two-stage sample splitting meta-algorithm that takes as input two arbitrary estimation algorithms: one for the target parameter and one for the nuisance parameter. We show that if the population risk satisfies a condition called Neyman orthogonality, the impact of the nuisance estimation error on the excess risk bound achieved by the meta-algorithm is of second order. Our theorem is agnostic to the particular algorithms used for the target and nuisance and only makes an assumption on their individual performance. This enables the use of a plethora of existing results from statistical learning and machine learning literature to give new guarantees for learning with a nuisance component. Moreover, by focusing on excess risk rather than parameter estimation, we can give guarantees under weaker assumptions than in previous works and accommodate the case where the target parameter belongs to a complex nonparametric class. We characterize conditions on the metric entropy such that oracle rates---rates of the same order as if we knew the nuisance parameter---are achieved. We also analyze the rates achieved by specific estimation algorithms such as variance-penalized empirical risk minimization, neural network estimation and sparse high-dimensional linear model estimation. We highlight the applicability of our results in four settings of central importance in the literature: 1) heterogeneous treatment effect estimation, 2) offline policy optimization, 3) domain adaptation, and 4) learning with missing data.

scholarly journals Forensic Feature-Comparison Expertise: Statistical Learning Facilitates Visual Comparison Performance

2020 ◽  
Author(s):  
Bethany Growns ◽  
Kristy Martire
Keyword(s):  
Statistical Learning ◽  
High Performance ◽  
Comparison Task ◽  
Domain Specific ◽  
Visual Evidence ◽  
Visual Comparison ◽  
Distributional Information ◽  
Distributional Learning ◽  
The Impact ◽  
The Relationship

Forensic feature-comparison examiners in select disciplines are more accurate than novices when comparing visual evidence samples. This paper examines a key cognitive mechanism that may contribute to this superior visual comparison performance: the ability to learn how often stimuli occur in the environment (distributional statistical learning). We examined the relation-ship between distributional learning and visual comparison performance, and the impact of training about the diagnosticity of distributional information in visual comparison tasks. We compared performance between novices given no training (uninformed novices; n = 32), accu-rate training (informed novices; n = 32) or inaccurate training (misinformed novices; n = 32) in Experiment 1; and between forensic examiners (n = 26), informed novices (n = 29) and unin-formed novices (n = 27) in Experiment 2. Across both experiments, forensic examiners and nov-ices performed significantly above chance in a visual comparison task where distributional learning was required for high performance. However, informed novices outperformed all par-ticipants and only their visual comparison performance was significantly associated with their distributional learning. It is likely that forensic examiners’ expertise is domain-specific and doesn’t generalise to novel visual comparison tasks. Nevertheless, diagnosticity training could be critical to the relationship between distributional learning and visual comparison performance.

scholarly journals Impact of Building Design Parameters on Daylighting Metrics Using an Analysis, Prediction, and Optimization Approach Based on Statistical Learning Technique

2019 ◽  
Vol 11 (5) ◽  
pp. 1474 ◽  
Author(s):  
Jaewook Lee ◽  
Mohamed Boubekri ◽  
Feng Liang
Keyword(s):  
Statistical Learning ◽  
Building Design ◽  
Design Parameters ◽  
Optimization Approach ◽  
Impact Performance ◽  
Design Variables ◽  
Design Attributes ◽  
Learning Technique ◽  
The Impact ◽  
The Relationship

Daylighting metrics are used to predict the daylight availability within a building and assess the performance of a fenestration solution. In this process, building design parameters are inseparable from these metrics; therefore, we need to know which parameters are truly important and how they impact performance. The purpose of this study is to explore the relationship between building design attributes and existing daylighting metrics based on a new methodology we are proposing. This methodology involves statistical learning. It is an emerging methodology that helps us to analyze a large quantity of output data and the impact of a large number of design variables. In particular, we can use these statistical methodologies to analyze which features are important, which ones are not, and the type of relationships they have. Using these techniques, statistical models may be created to predict daylighting metric values for different building types and design solutions. In this article we will outline how this methodology works, and analyze the building design features that have the strongest impact on daylighting performance.

The impact of phonological neighborhood density on typical and atypical emerging lexicons*

2013 ◽  
Vol 41 (3) ◽  
pp. 634-657 ◽  
Author(s):  
STEPHANIE F. STOKES
Keyword(s):  
Statistical Learning ◽  
Neighborhood Density ◽  
Phonological Representations ◽  
Phonological Neighborhood Density ◽  
Learning Mechanism ◽  
Late Talkers ◽  
Ambient Language ◽  
Current Article ◽  
Phonological Neighborhood ◽  
The Impact

ABSTRACTAccording to the Extended Statistical Learning account (ExSL; Stokes, Kern & dos Santos, 2012) late talkers (LTs) continue to use neighborhood density (ND) as a cue for word learning when their peers no longer use a density learning mechanism. In the current article, LTs expressive (active) lexicon ND values differed from those of their age-matched, but not language-matched, TD peers, a finding that provided support for the ExSL account. Stokes (2010) claimed that LTs had difficulty abstracting sparse words, but not dense, from the ambient language. If true, then LTs' receptive (passive), as well as active lexicons should be comprised of words of high ND. However, in the current research only active lexicons were of high ND. LTs' expressive lexicons may be small not because of an abstraction deficit, but because they are unable to develop sufficiently strong phonological representations to support word production.

Strengthening risk prediction using statistical learning in children with autism spectrum disorder

2018 ◽  
Vol 4 (3) ◽  
pp. 141-152 ◽  
Author(s):  
Tanu Tanu ◽  
Deepti Kakkar
Keyword(s):  
Statistical Learning ◽  
Risk Prediction ◽  
Autism Spectrum ◽  
Second Phase ◽  
Prediction Ability ◽  
Content Type ◽  
Risk Levels ◽  
Children With Asd ◽  
The Impact ◽  
Risk Knowledge

Purpose The purpose of this paper is to investigate the prediction ability in children with ASD in the risk-involving situations and compute the impact of statistical learning (SL) in strengthening their risk knowledge. The learning index and stability with time are also calculated by comparing their performance over three consecutive weekly sessions (session 1, session 2 and session 3). Design/methodology/approach Participants were presented with a series of images, showing simple and complex risk-involving situations, using the psychophysical experimental paradigm. The stimuli in the experiment were provided with different levels of difficulty in order to keep the legacy of the prediction and SL-based experiment intact. Findings The first phase of experimental work showed that children with ASD accurately discriminated the risk, although performed poorly as compared to neurotypical. The attenuated response in differentiating risk levels indicates that children with ASD have a poor and underdeveloped sense of risk. The second phase investigated their capability to extract the information from repetitive patterns and calculated SL stability value in time. The learning curve shows that SL is intact and stable with time (average session r=0.74) in children with ASD. Research limitations/implications The present work concludes that impaired action prediction could possibly be one of the factors underlying underdeveloped sense of risk in children with ASD. Their SL capability shows that risk knowledge can be strengthened in them. In future, the studies should investigate the impact of age and individual differences, by using knowledge from repetitive trials, on the learning rate and trajectories. Practical implications SL, being an integral part of different therapies, rehabilitation schemes and intervention systems, has the potential to enhance the cognitive and functional abilities of children with ASD. Originality/value Past studies have provided evidence regarding the work on the prediction ability in individuals with ASD. However, it is unclear whether the risk-involving/dangerous situations play any certain role to enhance the prediction ability in children with ASD. Also, there are limited studies predicting risk knowledge in them. Based on this, the current work has investigated the risk prediction in children with ASD.

Statistics, Structure, and Style in Music

2015 ◽  
Vol 33 (1) ◽  
pp. 20-31 ◽  
Author(s):  
Carol Lynne Krumhansl
Keyword(s):  
Statistical Learning ◽  
Music Cognition ◽  
Contextual Effects ◽  
Psychological Processes ◽  
Psychological Measurement ◽  
The Third ◽  
Musical Pitch ◽  
Cognitive Foundations ◽  
The Impact

This essay begins by reviewing issues in psychological measurement that motivated some of the research summarized in Cognitive Foundations of Musical Pitch (Krumhansl, 1990). These were challenges to geometrical models of similarity, asymmetrical measures of similarity, and contextual effects. It then considers the impact that statistical learning has had on research and theory about music cognition, suggesting this emphasis may have underestimated the importance of other psychological processes contributing to the experience of music. Finally, it discusses three problems with traditional analyses using unidimensional, sequential statistics. The first is that music is hierarchical, with important relationships between non-adjacent events. The second is that the dimensions of music, specifically, pitch and time, interact. The third is the assumption that probabilities remain constant throughout a composition. Rather, music contains artful deviations from normative probabilities contributing to the experience of tension and release.

Sign in / Sign up

Export Citation Format

Share Document