An inter-battery method of factor analysis

Tucker, Ledyard R.

doi:10.1007/bf02289009

Cited by 310 publications

(207 citation statements)

References 18 publications

Supporting

Mentioning

201

Contrasting

Unclassified

Order By: Relevance

“…However, the data are noisy, and the noise in the data may be inconsistent with the label which will cause poor performance when the data has low signal-noise ratio. The concept, which use factorized representation for information that can be shared among different views and information that cannot be shared between different views, has been applied to different frameworks with a long history [20,4,7]. We will adopt the same concept for supervised topic models.…”

Section: Modelsmentioning

confidence: 99%

How to Supervise Topic Models

Zhang

2015

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Supervised topic models are important machine learning tools which have been widely used in computer vision as well as in other domains. However, there is a gap in the understanding of the supervision impact on the model. In this paper, we present a thorough analysis on the behaviour of supervised topic models using Supervised Latent Dirichlet Allocation (SLDA) and propose two factorized supervised topic models, which factorize the topics into signal and noise. Experimental results on both synthetic data and real-world data for computer vision tasks show that supervision need to be boosted to be effective and factorized topic models are able to enhance the performance.

show abstract

Section: Modelsmentioning

confidence: 99%

How to Supervise Topic Models

Zhang

2015

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…The method is related to other multivariate analysis such as canonical correspondence analysis , redundancy analysis, and canonical correlation analysis (CANCOR) (Gittins, 1985). COIA is a generalization of the inter-battery analysis by Tucker (Tucker, 1958) which in turn is the first step of partial least squares (skuldsson, 1988). COIA is very similar to CANCOR in many aspects.…”

Section: Co-inertia Analysismentioning

confidence: 99%

Feature Matching using CO-Inertia Analysis for People Tracking

Simha

Chau

Brémond

2014

Proceedings of the 9th International Conference on Computer Vision Theory and Applications

View full text Add to dashboard Cite

Abstract:Robust object tracking is a challenging computer vision problem due to dynamic changes in object pose, illumination, appearance and occlusions. Tracking objects between frames requires accurate matching of their features. We investigate real time matching of mobile object features for frame to frame tracking. This paper presents a new feature matching approach between objects for tracking that incorporates one of the multivariate analysis method called Co-Inertia Analysis abbreviated as COIA. This approach is being introduced to compute the similarity between Histogram of Oriented Gradients (HOG) features of the tracked objects. Experiments conducted shows the effectiveness of this approach for mobile object feature tracking.

show abstract

“…PCA can also be used to process fully paired multimodal data (by stacking the data vectors), but this does not qualify as a multimodal technique in the sense of Definition 2, since the construction requires that all modalities are also present in future data. The truly multimodal counterpart to PCA is maximum covariance analysis (MCA) [31], which would be ideal for our purposes, except that it also requires fully paired data.…”

Section: Linear Weakly Paired Covariance Maximizationmentioning

confidence: 99%

Weakly-Paired Maximum Covariance Analysis for Multimodal Dimensionality Reduction and Transfer Learning

Lampert

Krömer

2010

Computer Vision – ECCV 2010

View full text Add to dashboard Cite

Abstract. We study the problem of multimodal dimensionality reduction assuming that data samples can be missing at training time, and not all data modalities may be present at application time. Maximum covariance analysis, as a generalization of PCA, has many desirable properties, but its application to practical problems is limited by its need for perfectly paired data. We overcome this limitation by a latent variable approach that allows working with weakly paired data and is still able to efficiently process large datasets using standard numerical routines. The resulting weakly paired maximum covariance analysis often finds better representations than alternative methods, as we show in two exemplary tasks: texture discrimination and transfer learning.

show abstract

An inter-battery method of factor analysis

Cited by 310 publications

References 18 publications

How to Supervise Topic Models

How to Supervise Topic Models

Feature Matching using CO-Inertia Analysis for People Tracking

Weakly-Paired Maximum Covariance Analysis for Multimodal Dimensionality Reduction and Transfer Learning

Contact Info

Product

Resources

About