Academic
Publications
The rate adapting poisson model for information retrieval and object recognition

The rate adapting poisson model for information retrieval and object recognition,10.1145/1143844.1143887,Peter V. Gehler,Alex D. Holub,Max Welling

The rate adapting poisson model for information retrieval and object recognition   (Citations: 23)
BibTex | RIS | RefWorks Download
Probabilistic modelling of text data in the bag- of-words representation has been dominated by directed graphical models such as pLSI, LDA, NMF, and discrete PCA. Recently, state of the art performance on visual object recognition has also been reported using variants of these mod- els. We introduce an alternative undirected graphical model suitable for modelling count data. This "Rate Adapting Poisson" (RAP) model is shown to generate superior dimension- ally reduced representations for subsequent re- trieval or classification. Models are trained us- ing contrastive divergence while inference of la- tent topical representations is efficiently achieved through a simple matrix multiplication.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...Several further developments of these undirected models [6, 17] show that they are competitive in terms of retrieval accuracy with their directed counterparts...
    • ...This model can be viewed as a variant of the Rate Adaptive Poisson model [6] that is easier to train and has a better way of dealing with documents of different lengths...
    • ...Results of [6] show that pLSA and LDA models do not generally outperform LSA and TF-IDF...

    Ruslan Salakhutdinovet al. Semantic hashing

    • ...A number of extensions and alternatives have been proposed to address these issues: latent Dirichlet allocation [4], undirected PLSI [28], correlated topic models [3], rate adapting Poisson models [7]; but they come at the price of an increased complexity, especially regarding the runtime cost for the learning algorithms...

    Jean-cédric Chappelieret al. PLSI: The True Fisher Kernel and beyond

    • ...A tf-idf vector space model and LSI [11] are two main baselines we will compare to. We already mentioned that pLSA [19] and LDA [3] both have scalability problems and are not reported to generally outperform LSA and TF-IDF [13]...

    Bing Baiet al. Supervised semantic indexing

    • ...Restricted Boltzmann Machines (RBMs) (Hinton et al., 2006; Smolensky, 1986) are neural network models for unsupervised learning, but have recently seen a lot of application as feature extraction methods for supervised learning algorithms (Salakhutdinov et al., 2007; Larochelle et al., 2007; Bengio et al., 2007; Gehler et al., 2006; Hinton et al., 2006; Hinton & Salakhutdinov, 2006)...

    Tijmen Tieleman. Training restricted Boltzmann machines using approximations to the lik...

    • ...image data (Gehler et al., 2006) or as a good initial training phase for deep neural network classifiers (Hinton, 2007)...
    • ...Gehler et al. (2006); Xing et al. (2005) have shown that the features learned by an RBM trained by ignoring the labeled targets can be useful for retrieving documents or classifying images of objects...

    Hugo Larochelleet al. Classification using discriminative restricted Boltzmann machines

Sort by: