Filters








835 Hits in 3.6 sec

Poisson Dependency Networks: Gradient Boosted Models for Multivariate Count Data

Fabian Hadiji, Alejandro Molina, Sriraam Natarajan, Kristian Kersting
2015 Machine Learning  
To ease the modeling of multivariate count data, we therefore introduce a novel family of Poisson graphical models, called Poisson Dependency Networks (PDNs).  ...  Although count data are increasingly ubiquitous, surprisingly little work has employed probabilistic graphical models for modeling count data.  ...  Acknowledgments The authors would like to thank Ying-Wooi Wan and Zhandong Liu for making the code for Poisson graphical models and for the simulated network recovery experiments from multivariate count  ... 
doi:10.1007/s10994-015-5506-z fatcat:ifotxlp3gncejoikpblx4ynaqu

Coresets for Dependency Networks [article]

Alejandro Molina, Alexander Munteanu, Kristian Kersting
2017 arXiv   pre-print
In this paper, we show how to construct coresets -compressed data sets which can be used as proxy for the original data and have provably bounded worst case error- for Gaussian dependency networks (DNs  ...  Despite this worst-case result, we will provide an argument why our coreset construction for DNs can still work well in practice on count data.  ...  structure learning using local estimtatiors or gradient tree boosting.  ... 
arXiv:1710.03285v2 fatcat:u2ejck5evbaaxkvg6jloxwq2d4

Predictive Analytics of Insurance Claims Using Multivariate Decision Trees

Zhiyu Quan, Emiliano A. Valdez
2018 Social Science Research Network  
Innovations to the original methods, such as random forests and gradient boosting, have further improved the capabilities of using decision trees as a predictive model.  ...  In addition, the extension of using decision trees with multivariate response variables started to develop and it is the purpose of this paper to apply multivariate tree models to insurance claims data  ...  The data used in this paper was provided by Gee Lee and Edward W. (Jed) Frees of the University of Wisconsin in Madison; we extend our appreciation to them for allowing us to use the data.  ... 
doi:10.2139/ssrn.3216135 fatcat:ruzpxccmcnhnpnb45wblmwkzpi

Supervised Multiscale Dimension Reduction for Spatial Interaction Networks [article]

Shaobo Han, David B. Dunson
2019 arXiv   pre-print
We introduce a multiscale supervised dimension reduction method for SPatial Interaction Network (SPIN) data, which consist of a collection of spatially coordinated interactions.  ...  We consider an inverse Poisson regression model and propose a new multiscale generalized double Pareto prior, which is induced via a tree-structured parameter expansion scheme.  ...  effect slope parameter for the m simple Poisson mixed regression model (Hall et al., 2011b) .  ... 
arXiv:1901.00172v3 fatcat:yjypvgd4anaaxmvm55gx36s6vm

Forecasting Emergency Calls with a Poisson Neural Network-based Assemble Model

Hongyun Huang, Mingyue Jiang, Zuohua Ding, Mengchu Zhou
2019 IEEE Access  
INDEX TERMS Accident forecasting, combined model, Poisson distribution, Poisson neural network, residual model.  ...  In this paper, a combined model, which consists of two parts, is proposed. The first part is a Poisson neural network model (PNN).  ...  POISSON NEURAL NETWORK We combine a Poisson regression model and NN model to produce a PNN model.  ... 
doi:10.1109/access.2019.2896887 fatcat:uixrdkan7neabdsz2h6edvqgoa

Joint learning of multiple gene networks from single-cell gene expression data

Nuosi Wu, Fu Yin, Le Ou-Yang, Zexuan Zhu, Weixin Xie
2020 Computational and Structural Biotechnology Journal  
Our model can deal with non-Gaussian data with missing values, and identify the common and unique network structures of multiple cell subgroups, which is suitable for scRNA-seq data.  ...  In this study, we introduce a novel joint Gaussian copula graphical model (JGCGM) to jointly estimate multiple gene networks for multiple cell subgroups from scRNA-seq data.  ...  Data generation 3.1.1. Poisson graphical models for count data In order to verify the effectiveness of our proposed model, we generate simulation data with known network structures.  ... 
doi:10.1016/j.csbj.2020.09.004 pmid:33033579 pmcid:PMC7527714 fatcat:broayde7mndq3j5qhdrqhttefi

Spatially-Adaptive Reconstruction in Computed Tomography Using Neural Networks

David Boublil, Michael Elad, Joseph Shtok, Michael Zibulevsky
2015 IEEE Transactions on Medical Imaging  
We propose a supervised machine learning approach for boosting existing signal and image recovery methods and demonstrate its efficacy on example of image reconstruction in computed tomography.  ...  The fusion of the images is performed by feed-forward neural network trained on a set of known examples.  ...  In practice, these photon count measurements are degraded by stochastic noise, typically modeled as instances of Poisson random variables.  ... 
doi:10.1109/tmi.2015.2401131 pmid:25675453 fatcat:h6cpdkwqd5g4jb6swj5ye6ochq

Random Projection in Deep Neural Networks [article]

Piotr Iwo Wójcik
2018 arXiv   pre-print
The second area where the application of RP techniques can be beneficial for training deep models is weight initialization.  ...  We focus on two areas where, as we have found, employing RP techniques can improve deep models: training neural networks on high-dimensional data and initialization of network parameters.  ...  The reference networks, on the other hand, used the constrained Poisson model, which is tailored specifically for word count data.  ... 
arXiv:1812.09489v1 fatcat:ioaksqy4v5h2fcn2c7tjshteya

Statistical inference of regulatory networks for circadian regulation

Andrej Aderhold, Dirk Husmeier, Marco Grzegorczyk
2014 Statistical Applications in Genetics and Molecular Biology  
of systematically missing values related to unknown protein concentrations and mRNA transcription rates, it investigates the dependence of the performance on the network topology and the degree of recurrency  ...  We assess the accuracy of various state-of-the-art statistics and machine learning methods for reconstructing gene and protein regulatory networks in the context of circadian regulation.  ...  Anne Smith for helpful discussions.  ... 
doi:10.1515/sagmb-2013-0051 pmid:24864301 fatcat:6po5swfc25a6lisphtyrtyamwy

Spatially-Adaptive Reconstruction in Computed Tomography using Neural Networks [article]

Joseph Shtok, Michael Zibulevsky, Michael Elad
2013 arXiv   pre-print
We propose a supervised machine learning approach for boosting existing signal and image recovery methods and demonstrate its efficacy on example of image reconstruction in computed tomography.  ...  The fusion of the images is performed by feed-forward neural network trained on a set of known examples.  ...  In practice, these photon count measurements are degraded by stochastic noise, typically modeled as instances of Poisson random variables.  ... 
arXiv:1311.7251v1 fatcat:o2bkjflp6ngttcx3gzreju3ayq

Conditional Sum-Product Networks: Imposing Structure on Deep Probabilistic Architectures [article]

Xiaoting Shao, Alejandro Molina, Antonio Vergari, Karl Stelzner, Robert Peharz, Thomas Liebig, Kristian Kersting
2019 arXiv   pre-print
Therefore, we introduce conditional SPNs (CSPNs), conditional density estimators for multivariate and potentially hybrid domains which allow harnessing the expressive power of neural networks while still  ...  Furthermore, they can successfully be employed as building blocks for structured probabilistic models, such as autoregressive image models.  ...  Kristian Kersting also acknowledges the support of the Rhine-Main Universities Network for "Deep Continuous-Discrete Machine Learning" (DeCoDeML).  ... 
arXiv:1905.08550v2 fatcat:kdhsmcphv5gspcetvugr5vhr7a

Compressed Self-Attention for Deep Metric Learning with Low-Rank Approximation

Ziye Chen, Mingming Gong, Lingjuan Ge, Bo Du
2020 Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence  
In this paper, we apply self-attention (SA) mechanism to boost the performance of deep metric learning.  ...  However, due to the pairwise similarity measurement, the cost of storing and manipulating the complete attention maps makes it infeasible for large inputs.  ...  Assuming that dataset of V -dimensional sequentially observed multivariate count data x 1 , ..., x t are represented as a V × T count matrix X.  ... 
doi:10.24963/ijcai.2020/281 dblp:conf/ijcai/Chen0LZZ20 fatcat:vafvkl6o4zcuno7cjt427dyuuu

Manifold Oblique Random Forests: Towards Closing the Gap on Convolutional Deep Networks [article]

Ronan Perry, Adam Li, Chester Huynh, Tyler M. Tomita, Ronak Mehta, Jesus Arroyo, Jesse Patsolic, Benjamin Falk, Joshua T. Vogelstein
2021 arXiv   pre-print
Decision forests (Forests), in particular random forests and gradient boosting trees, have demonstrated state-of-the-art accuracy compared to other methods in many supervised learning scenarios.  ...  However, in structured data lying on a manifold (such as images, text, and speech) deep networks (Networks), specifically convolutional deep networks (ConvNets), tend to outperform Forests.  ...  For each sampled image, k ∼ Poisson(λ = 10) bars were distributed among the rows or columns, depending on the class.  ... 
arXiv:1909.11799v4 fatcat:xgpplctf25gmlon2o4jolv5sp4

Predicting Spatial Crime Occurrences through an Efficient Ensemble-Learning Model

Yasmine Lamari, Bartol Freskura, Anass Abdessamad, Sarah Eichberg, Simon de Bonviller
2020 ISPRS International Journal of Geo-Information  
The gradient boosting model was found to yield the most accurate predictions for violent crimes, property crimes, motor vehicle thefts, vandalism, and the total count of crimes.  ...  Such data are published periodically for the entire United States.  ...  Lamari, Bartol Freskura, and Anass Abdessamad; Investigation, Bartol Freskura and Anass Abdessamad; Resources, Simon de Bonviller, Yasmine Lamari, Anass Abdessamad, Sarah Eichberg, and Bartol Freskura; Data  ... 
doi:10.3390/ijgi9110645 fatcat:osco5gsktrgw5mdcw76gnqmrqu

Agent-Based Model Calibration Using Machine Learning Surrogates

Francesco Lamperti, Amir Sani
2017 Social Science Research Network  
Abstract Taking agent-based models (ABM) closer to the data is an open challenge.  ...  Performance is evaluated against a relatively large out-of-sample set of parameter combinations, while employing different user-defined statistical tests for output analysis.  ...  We would also like to thank Daniele Giachini, Mattia Guerini, Matteo Sostero and Baláz Kégl for their comments.  ... 
doi:10.2139/ssrn.2943297 fatcat:idldj43v5zbbjfacdmtb4ipdja
« Previous Showing results 1 — 15 out of 835 results