Filters








21,350 Hits in 7.4 sec

Beyond Proportional Fairness: A Resource Biasing Framework for Shaping Throughput Profiles in Multihop Wireless Networks

S. Singh, U. Madhow, E. M. Belding
2008 2008 Proceedings IEEE INFOCOM - The 27th Conference on Computer Communications  
We present an analytical model that gives insight into the impact of a particular resource allocation strategy on network performance, in a manner that captures the effect of finite network size and spatial  ...  Throughput performance of multihop wireless networks is governed by how the network's transport capacity (in bit-meters per second) is partitioned among different network flows.  ...  ACKNOWLEDGMENTS This work was supported in part by NSF Grant CCF-0431205, ONR Grant N00014-06-1-0066, NSF Career Award CNS-0347886 and NSF NeTS Award CNS-0435527.  ... 
doi:10.1109/infocom.2007.309 fatcat:h33h63qebnfnphonqeibbrqphy

Beyond Proportional Fairness: A Resource Biasing Framework for Shaping Throughput Profiles in Multihop Wireless Networks

S. Singh, U. Madhow, E. M. Belding
2008 IEEE INFOCOM 2008 - The 27th Conference on Computer Communications  
We present an analytical model that gives insight into the impact of a particular resource allocation strategy on network performance, in a manner that captures the effect of finite network size and spatial  ...  Throughput performance of multihop wireless networks is governed by how the network's transport capacity (in bit-meters per second) is partitioned among different network flows.  ...  ACKNOWLEDGMENTS This work was supported in part by NSF Grant CCF-0431205, ONR Grant N00014-06-1-0066, NSF Career Award CNS-0347886 and NSF NeTS Award CNS-0435527.  ... 
doi:10.1109/infocom.2008.309 dblp:conf/infocom/SinghMB08 fatcat:kdvyslx5nngfxnc47m4socckmq

Shaping Throughput Profiles in Multihop Wireless Networks: A Resource-Biasing Approach

Sumit Singh, U. Madhow, E. M. Belding
2012 IEEE Transactions on Mobile Computing  
We present an analytical model that offers insight into the impact of a particular resource allocation strategy on network performance, taking into account finite network size and spatial traffic patterns  ...  Specifically, mixing strongly biased allocations with fairer allocations leads to efficient network utilization as well as a superior trade-off between flow throughput and fairness.  ...  We now employ simulations to obtain flow throughput profiles, and to evaluate the benefits of mixed-bias strategies relative to the individual strategies in the mixture, as well as to proportional fairness  ... 
doi:10.1109/tmc.2011.63 fatcat:gsr67xle2zbalgyvdp2tz5oexe

Statistical Learning Algorithms Applied to Automobile Insurance Ratemaking [chapter]

C. Dugas, Y. Bengio, N. Chapados, P. Vincent, G. Denoncourt, C. Fournier
2003 Intelligent and Other Computational Techniques in Insurance  
The main numerical result is a statistically significant reduction in the out-of-sample meansquared error using the neural network model and our ability to substantially reduce the median premium by charging  ...  We analyzed the performance of several models within five broad categories: linear regressions, generalized linear models, decision trees, neural networks and support vector machines.  ...  At the same time, the Mixture model achieves better prediction accuracy, as measured by the Mean-Squared Error (MSE) of the respective models, all the while remaining fair to customers in all categories  ... 
doi:10.1142/9789812794246_0004 fatcat:lwx3qtshqjdkpjevnm27ivujiu

Fair Generative Modeling via Weak Supervision [article]

Kristy Choi, Aditya Grover, Trisha Singh, Rui Shu, Stefano Ermon
2020 arXiv   pre-print
Using this supplementary dataset, we detect the bias in existing datasets via a density ratio technique and learn generative models which efficiently achieve the twin goals of: 1) data efficiency by using  ...  training examples from both biased and reference datasets for learning; and 2) data generation close in distribution to the reference dataset at test time.  ...  KC is supported by the NSF GRFP, Qualcomm Innovation Fellowship, and Stanford Graduate Fellowship, and AG is supported by the MSR Ph.D. fellowship, Stanford Data Science scholarship, and Lieberman fellowship  ... 
arXiv:1910.12008v2 fatcat:zc46etpoezafneo6yk7a644nbq

Journal of computing in civil engineering

Fred Collopy
1996 International Journal of Forecasting  
The objective of this study was to develop an improved and more effective dynamic modulus regression model for mixtures in Costa Rica using Neural Networks.  ...  Results indicated that the new and improved model based on neural networks (E* NN-model) not only met the model adequacy checking criteria but also exhibited the best goodness of fit parameters and the  ...  The results also allowed the correction of the bias between predicted and measured E* by means of statistical calibration.  ... 
doi:10.1016/0169-2070(96)88200-8 fatcat:hx62pu5qabcgtlxmpacu7m7nwy

Neural Network Based Model to Estimate Dynamic Modulus E* for Mixtures in Costa Rica

Fabricio Leiva-Villacorta, Adriana Vargas-Nordcbeck
2019 Journal of Soft Computing in Civil Engineering  
The objective of this study was to develop an improved and more effective dynamic modulus regression model for mixtures in Costa Rica using Neural Networks.  ...  Results indicated that the new and improved model based on neural networks (E* NN-model) not only met the model adequacy checking criteria but also exhibited the best goodness of fit parameters and the  ...  The results also allowed the correction of the bias between predicted and measured E* by means of statistical calibration.  ... 
doi:10.22115/scce.2019.188006.1110 doaj:3d393bc55bc0433b8714d47c14ff561b fatcat:tgweloifsfhtvb3rwte5lol7xq

Bias and Debias in Recommender System: A Survey and Future Directions [article]

Jiawei Chen, Hande Dong, Xiang Wang, Fuli Feng, Meng Wang, Xiangnan He
2021 arXiv   pre-print
This makes various biases widely exist in the data, including but not limited to selection bias, position bias, exposure bias, and popularity bias.  ...  In this paper, we first summarize seven types of biases in recommendation, along with their definitions and characteristics.  ...  of seven types of biases in recommendation and the bias amplification in loop.  ... 
arXiv:2010.03240v2 fatcat:6fticc3otndsra2whs5e4nrdpi

Proposing an Interactive Audit Pipeline for Visual Privacy Research [article]

Jasmine DeHart, Chenguang Xu, Lisa Egede, Christan Grant
2021 arXiv   pre-print
The continued use of biased datasets and processes will adversely damage communities and increase the cost of fixing the problem later.  ...  In an ideal world, deployed machine learning models will enhance our society. We hope that those models will provide unbiased and ethical decisions that will benefit everyone.  ...  Acknowledgments The researchers are partially supported by awards from the Department of Defense SMART Scholarship and the National Science Foundation under Grant No. #1952181.  ... 
arXiv:2111.03984v2 fatcat:chrnfyevfbc5ljzxdxvlpuxody

Deep Neural Network for Analysis of DNA Methylation Data [article]

Hong Yu, Zhanyu Ma
2020 arXiv   pre-print
Experiments show these features perform best in breast cancer DNA methylation data cluster analysis, comparing with some state-of-the-art methods.  ...  In order to explicitly capture the properties of the data, we design a deep neural network, which composes of several stacked binary restricted Boltzmann machines, to learn the low dimensional deep features  ...  (c), In order to make fair comparisons, we also applied the k-means, the Gaussian mixture model (GMM), and the SOM method to cluster the features extracted by PCA, NMF, and DNN methods, respectively.  ... 
arXiv:1808.01359v2 fatcat:ewerszc5afgrlfskuw4i26na4e

Regularization Strategy for Point Cloud via Rigidly Mixed Sample [article]

Dogyoon Lee, Jaeha Lee, Junhyeop Lee, Hyeongmin Lee, Minhyeok Lee, Sungmin Woo, Sangyoun Lee
2021 arXiv   pre-print
Data augmentation is an effective regularization strategy to alleviate the overfitting, which is an inherent drawback of the deep neural networks.  ...  We also analyzed various combinations of data augmentations including RSMix with single and multi-view evaluations, based on abundant ablation studies.  ...  All the results reveal that RSMix improved the network accuracies regardless of network type or evaluation methods, verifying the effectiveness of our shape-preserved mixture approach with significant  ... 
arXiv:2102.01929v3 fatcat:nrifgdo73zdffkegq255mfzwny

The revelation effect for autobiographical memory: A mixture-model analysis

Daniel M. Bernstein, Michael E. Rudd, Edgar Erdfelder, Ryan Godfrey, Elizabeth F. Loftus
2009 Psychonomic Bulletin & Review  
Our analysis reveals that unscrambling a key word or an unrelated word affects response bias and discriminability in autobiographical memory tests in ways that are very similar to those that have been  ...  We analyze our data using a new signal detection mixture distribution model which does not require that the researcher knows the veracity of memory judgments a priori.  ...  Table 2 . 2 Mean parameter estimates (and standard errors of means) for the signal detection mixture model (Experiments 1 and 2). Experiment 1 Familiarity Means Familiarity Std.  ... 
doi:10.3758/pbr.16.3.463 pmid:19451369 fatcat:3mq7jagbg5e7jdkyyh3efzdd6u

Recurrent Mixture Density Network for Spatiotemporal Visual Attention [article]

Loris Bazzani and Hugo Larochelle and Lorenzo Torresani
2017 arXiv   pre-print
Time consistency in videos is modeled hierarchically by: 1) deep 3D convolutional features to represent spatial and short-term time relations and 2) a long short-term memory network on top that aggregates  ...  We model visual attention with a mixture of Gaussians at each frame. This distribution is used to express the probability of saliency for each pixel.  ...  We are grateful to Stefan Mathe for explaining the format of the eyetracking data and the protocol of the Hollywood2 experiment. This work was funded in part by NSF award CNS-1205521.  ... 
arXiv:1603.08199v4 fatcat:vuweqrzfgncnpnlojxgsgrdqca

TabFairGAN: Fair Tabular Data Generation with Generative Adversarial Networks [article]

Amirarsalan Rajabi, Ozlem Ozmen Garibay
2021 arXiv   pre-print
Also, in the constrained case in which the first phase of training is followed by the second phase, we train the network and test it on four datasets studied in the fairness literature and compare our  ...  Comparing to other studies utilizing GANs for fair data generation, our model is comparably more stable by using only one critic, and also by avoiding major problems of original GAN model, such as mode-dropping  ...  For each model, we train five times and report the means and standard deviations of evaluation results in Table II .  ... 
arXiv:2109.00666v1 fatcat:lfwfhxixjbbfdldqrzdfzq3kq4

Photometric redshift estimation via deep learning

A. D'Isanto, K. L. Polsterer
2018 Astronomy and Astrophysics  
The estimates are expressed as Gaussian mixture models representing the probability density functions (PDFs) in the redshift space.  ...  A modified version of a deep convolutional network was combined with a mixture density network.  ...  Funding for SDSS-III has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, and the U.S. Department of Energy Office of Science.  ... 
doi:10.1051/0004-6361/201731326 fatcat:7nlodhpn6balhli6hp63f72ieu
« Previous Showing results 1 — 15 out of 21,350 results