Filters








10,723 Hits in 4.8 sec

Learning from Censored and Dependent Data: The case of Linear Dynamics [article]

Orestis Plevrakis
2021
In light of recent developments on learning linear dynamical systems (LDSs), and on censored statistics with independent data, we revisit the decades-old problem of learning an LDS, from censored observations  ...  Observations from dynamical systems often exhibit irregularities, such as censoring, where values are recorded only if they fall within a certain range.  ...  Introduction System identification is the problem of learning the evolution equations of a dynamical system from data.  ... 
doi:10.48550/arxiv.2104.05087 fatcat:k5xpwioh4jfjbdgr56f55hvwem

Constructing Stabilized Dynamic Treatment Regimes for Censored Data [article]

Ying-Qi Zhao and Ruoqing Zhu and Guanhua Chen and Yingye Zheng
2019 arXiv   pre-print
To address this challenge, we propose two novel methods, censored shared-Q-learning and censored shared-O-learning.  ...  The estimation of stabilized dynamic treatment regimes becomes more complicated when the clinical outcome of interest is a survival time subject to censoring.  ...  In this paper, we let the initial values depend on the estimates from censored Q-learning (Goldberg and Kosorok, 2012) , where the parameters are not shared.  ... 
arXiv:1808.01332v2 fatcat:evp5popdijhszgr66ekgqgwxbm

Q-learning with censored data

Yair Goldberg, Michael R. Kosorok
2012 Annals of Statistics  
We present a novel Q-learning algorithm that is adjusted for censored data and allows a flexible number of stages.  ...  We simulate a multistage clinical trial with flexible number of stages and apply the proposed censored-Q-learning algorithm to find individualized treatment regimens.  ...  The authors are grateful to the anonymous reviewers and the Associate Editor for their helpful suggestions and comments. SUPPLEMENTARY MATERIAL  ... 
doi:10.1214/12-aos968 pmid:22754029 pmcid:PMC3385950 fatcat:ojpdzgxgrng77bpqxfwnvmw4sa

Predicting Winning Price in Real Time Bidding with Censored Data

Wush Chi-Hsuan Wu, Mi-Yen Yeh, Ming-Syan Chen
2015 Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '15  
We propose to leverage the machine learning and statistical methods to train the winning price model from the bidding history.  ...  Note , however, the assumption of censored regression does not hold on the real RTB data.  ...  Because the censored regression learns from more data compared to the linear regression, which learns from the winning bids only, the estimator of the censored regression β clm is supposed to have a smaller  ... 
doi:10.1145/2783258.2783276 dblp:conf/kdd/WuYC15 fatcat:dix7w3an6nbcbfuvuxxp3aqnbe

Doubly robust learning for estimating individualized treatment with censored data

Y. Q. Zhao, D. Zeng, E. B. Laber, R. Song, M. Yuan, M. R. Kosorok
2014 Biometrika  
We illustrate the proposed methods using simulation study and data from a Phase III clinical trial on non-small cell lung cancer.  ...  We develop nonparametric methods for estimating an optimal individualized treatment rule in the presence of censored data.  ...  A linear basis is applied for model fitting in Q-learning. Linear kernels were used for both the implementation of inverse censoring weighted and doubly robust outcome weighted learning.  ... 
doi:10.1093/biomet/asu050 pmid:25937641 pmcid:PMC4414056 fatcat:z74jtsizlnbbzmtjhlr3uggiga

Identification of Panel Data Models with Endogenous Censoring

Shakeeb Khan, Maria Ponomareva, Elie T. Tamer
2011 Social Science Research Network  
This paper analyzes the identification question in censored panel data models, where the censoring can depend on both observable and unobservable variables in arbitrary ways.  ...  We also extend our results to dynamic versions of the censored panel models in which we consider lagged observed, latent dependent variables and lagged censoring indicator variables as regressors.  ...  Our main contribution is to provide the tightest sets on the parameter of interest that we can learn from data at hand under two sets of assumptions.  ... 
doi:10.2139/ssrn.1831402 fatcat:i7ppvoamknbuhi2z44dscmui7u

Tree based weighted learning for estimating individualized treatment rules with censored data [article]

Yifan Cui, Ruoqing Zhu, Michael Kosorok
2017 arXiv   pre-print
In this paper, we extend the outcome weighted learning to right censored survival data without requiring either an inverse probability of censoring weighting or a semiparametric modeling of the censoring  ...  [zhao2012estimating] and [zhang2012robust] proposed outcome weighted learning to estimate individualized treatment rules directly through maximizing the expected outcome without modeling the response directly  ...  We thank the editor, associated editor, and reviewers for helpful comments which led to an improved manuscript.  ... 
arXiv:1707.09632v2 fatcat:gvqqkqul35bb7f2phg45b6v7ym

Identification of panel data models with endogenous censoring

Shakeeb Khan, Maria Ponomareva, Elie Tamer
2016 Journal of Econometrics  
We study inference on parameters in censored panel data models, where the censoring can depend on both observable and unobservable variables in arbitrary ways.  ...  Also, we also show how our results extend to empirically interesting dynamic versions of the model with both lagged observed outcomes, and lagged indicators.  ...  Our main contribution is to provide the tightest sets on the parameter of interest that we can learn from data at hand under two sets of assumptions.  ... 
doi:10.1016/j.jeconom.2016.01.010 fatcat:w7v6phqbgnbcxfvvsayh3igdti

Estimating Dynamic Signals From Trial Data With Censored Values

Ali Yousefi, Darin D. Dougherty, Emad N. Eskandar, Alik S. Widge, Uri T. Eden
2017 Computational Psychiatry  
We showed that special cases of this framework can incorporate the three most common approaches to censored observations: ignoring trials with censored data, imputing the censored data values, or using  ...  We compared the performances of these methods in a simulation study and provide recommendations of approaches to use, based on the expected amount of censored data in an experiment.  ...  The views, opinions, and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S.  ... 
doi:10.1162/cpsy_a_00003 pmid:29601047 pmcid:PMC5774187 fatcat:fjdwpgwdlff5bafoi66p34e5fi

SurvBenchmark: comprehensive benchmarking study of survival analysis methods using both omics data and clinical data [article]

Yunwei Zhang, Germaine Wong, Graham Mann, Samuel Muller, Jean Yee Hwa Yang
2021 bioRxiv   pre-print
Survival analysis is a branch of statistics that deals with both, the tracking of time and of the survival status simultaneously as the dependent response.  ...  The key result is that there is no one size fits all solution for any of the criteria and any of the methods. Some methods with a high C-index suffer from computational exhaustion and instability.  ...  Acknowledgements The authors thank all their colleagues, particularly at The University of Sydney, Sydney Precision Bioinformatics Alliance and Charles Perkins Centre for their support and intellectual  ... 
doi:10.1101/2021.07.11.451967 fatcat:bdr3cuib6nflhmz2apaz2dtpye

Boost-R: Gradient Boosted Trees for Recurrence Data [article]

Xiao Liu, Rong Pan
2021 arXiv   pre-print
The sum of these functions, from multiple trees, yields the ensemble estimator of the cumulative intensity.  ...  This paper investigates an additive-tree-based approach, known as Boost-R (Boosting for Recurrence Data), for recurrent event data with both static and dynamic features.  ...  Conclusions This paper proposed an additive-tree-based statistical learning approach, known as Boost-R (Boosting for Recurrence Data), for modeling recurrent event data with both static and dynamic feature  ... 
arXiv:2107.08784v1 fatcat:rv27t52psvb27nqqrr4e3ybqdq

Bayesian network data imputation with application to survival tree analysis

Paola M.V. Rancoita, Marco Zaffalon, Emanuele Zucca, Francesco Bertoni, Cassio P. de Campos
2016 Computational Statistics & Data Analysis  
Acknowledgements The work of C. P. de Campos has been mostly done while he was affiliated with the Dalle Molle Institute for Artificial Intelligence and the Institute of Oncology Research.  ...  The work partially supported by a research grant from the Ente Ospedaliero Cantonale (EOC), Bellinzona, Switzerland; Oncosuisse (OCS-02034-02-2007, OCS-1939-8-2006; Swiss NSF grants Nos. 200021 146606  ...  The test was not significant for BS in the case of model A, with uncensored data and 15%-25% of missing data and with 25% of censored data and 15% of missing data (this latter case was also the only one  ... 
doi:10.1016/j.csda.2014.12.008 fatcat:fir54ibgtvgabdcy3ah7n6mzku

Game Data Mining Competition on Churn Prediction and Survival Analysis using Commercial Game Log Data

Eunjo Lee, Yoonjae Jang, Du-Mim Yoon, JiHoon Jeon, Sung-il Yang, SangKwang Lee, Dae-Wook Kim, Pei Pei Chen, Anna Guitart, Paul Bertens, Africa Perianez, Fabian Hadiji (+5 others)
2019 IEEE Transactions on Games  
The results of the competition revealed that highly ranked competitors used deep learning, tree boosting, and linear regression.  ...  The reluctance of these companies to make data publicly available limits the wide use and development of data mining techniques and artificial intelligence research specific to the game industry.  ...  This work is partially supported by the European Commission under grant agreement number 731900 -ENVISAGE.  ... 
doi:10.1109/tg.2018.2888863 fatcat:ifaukda6arhsjiz65jaasfjlly

On the Hardness of Inventory Management with Censored Demand Data [article]

Gábor Lugosi, Mihalis G. Markakis, Gergely Neu
2017 arXiv   pre-print
Through this result, we derive an important insight: the benefit from "information stalking" as well as the cost of censoring are both negligible in this dynamic learning problem, at least with respect  ...  We consider a repeated newsvendor problem where the inventory manager has no prior information about the demand, and can access only censored/sales data.  ...  To the best of our knowledge, this setting has not been studied before from the angle of demand learning via censored data, and from a nonstochastic viewpoint.  ... 
arXiv:1710.05739v1 fatcat:4r54au52wnevpfs5c7vxj5n4rm

Mislearning from Censored Data: The Gambler's Fallacy and Other Correlational Mistakes in Optimal-Stopping Problems [article]

Kevin He
2021 arXiv   pre-print
They are uncertain about the underlying distribution and learn its parameters from predecessors.  ...  I also analyze how other misperceptions of intertemporal correlation interact with endogenous data censoring.  ...  The degree of belief in this fictitious variation both depends on the severity of history censoring (as the amount of "noise" inferred depends on the kind of data) and influences the agents' stopping strategy  ... 
arXiv:1803.08170v6 fatcat:qqjpqazipbcv7c42wfv66yz6qe
« Previous Showing results 1 — 15 out of 10,723 results