scholarly journals Locally Linear Models on Face Appearance Manifolds with Application to Dual-Subspace Based Classification

2006 ◽  
Author(s):  
Wei Fan ◽  
Dit-Yan Yeung
Keyword(s):  
Linear Models ◽  
Dual Subspace ◽  
Face Appearance ◽  
Locally Linear

Constructive Incremental Learning from Only Local Information

1998 ◽  
Vol 10 (8) ◽  
pp. 2047-2084 ◽  
Author(s):  
Stefan Schaal ◽  
Christopher G. Atkeson
Keyword(s):  
Linear Model ◽  
Incremental Learning ◽  
Linear Models ◽  
Expert Knowledge ◽  
Receptive Fields ◽  
Learning System ◽  
Independent Learning ◽  
Independent Expert ◽  
Learning Capabilities ◽  
Locally Linear

We introduce a constructive, incremental learning system for regression problems that models data by means of spatially localized linear models. In contrast to other approaches, the size and shape of the receptive field of each locally linear model, as well as the parameters of the locally linear model itself, are learned independently, that is, without the need for competition or any other kind of communication. Independent learning is accomplished by incrementally minimizing a weighted local cross-validation error. As a result, we obtain a learning system that can allocate resources as needed while dealing with the bias-variance dilemma in a principled way. The spatial localization of the linear models increases robustness toward negative interference. Our learning system can be interpreted as a nonparametric adaptive bandwidth smoother, as a mixture of experts where the experts are trained in isolation, and as a learning system that profits from combining independent expert knowledge on the same problem. This article illustrates the potential learning capabilities of purely local learning and offers an interesting and powerful approach to learning with receptive fields.

scholarly journals Tangent-based manifold approximation with locally linear models

2014 ◽  
Vol 104 ◽  
pp. 232-247 ◽  
Author(s):  
Sofia Karygianni ◽  
Pascal Frossard
Keyword(s):  
Linear Models ◽  
Locally Linear

A Note on Effects of Digital Approximation in Linear Models

1974 ◽  
Vol 3 (9) ◽  
pp. 893-897
Author(s):  
Gerald McWilliams ◽  
James Poirot†
Keyword(s):  
Linear Models

Effects of Momentary Affect on Satisfaction Judgments

2020 ◽  
Vol 41 (2) ◽  
pp. 61-67
Author(s):  
Marko Tončić ◽  
Petra Anić
Keyword(s):  
Linear Models ◽  
Time Frame ◽  
Hierarchical Linear Models ◽  
Curvilinear Relationship ◽  
Positive And Negative Affect ◽  
Transient Events ◽  
Time Frames ◽  
Events Data ◽  
Level Satisfaction ◽  
Source Of Information

Abstract. This study aims to examine the effect of affect on satisfaction, both at the between- and the within-person level for momentary assessments. Affect is regarded as an important source of information for life satisfaction judgments. This affective effect on satisfaction is well established at the dispositional level, while at the within-person level it is heavily under-researched. This is true especially for momentary assessments. In this experience sampling study both mood and satisfaction scales were administered five times a day for 7 days via hand-held devices ( N = 74 with 2,122 assessments). Several hierarchical linear models were fitted to the data. Even though the amount of between-person variance was relatively low, both positive and negative affect had substantial effects on momentary satisfaction on the between- and the within-person level as well. The within-person effects of affect on satisfaction appear to be more pronounced than the between-person ones. At the momentary level, the amount of between-person variance is lower than in studies with longer time-frames. The affect-related effects on satisfaction possibly have a curvilinear relationship with the time-frame used, increasing in intensity up to a point and then decreasing again. Such a relationship suggests that, at the momentary level, satisfaction might behave in a more stochastic manner, allowing for transient events/data which are not necessarily affect-related to affect it.

Sign in / Sign up

Export Citation Format

Share Document