Filters








74,101 Hits in 6.9 sec

On-Line Maximum Likelihood Prediction with Respect to General Loss Functions

Kenji Yamanishi
1997 Journal of computer and system sciences (Print)  
This paper introduces a new family of deterministic and stochastic on-line prediction algorithms which work with respect to general loss functions and analyzes their behavior in terms of expected loss  ...  For all the cases, we derive upper bounds on the expected instantaneous or cumulative losses for the algorithms with respect to a large family of loss functions satisfying the constraint introduced by  ...  ACKNOWLEDGMENTS The author thanks Leonid Gurvits for helpful discussion and Jason Catlett for reading an earlier draft of this paper.  ... 
doi:10.1006/jcss.1997.1503 fatcat:opxw2qltffgtxiktkxp6e5apqu

On-line maximum likelihood prediction with respect to general loss functions [chapter]

Kenji Yamanishi
1995 Lecture Notes in Computer Science  
This paper introduces a new family of deterministic and stochastic on-line prediction algorithms which work with respect to general loss functions and analyzes their behavior in terms of expected loss  ...  For all the cases, we derive upper bounds on the expected instantaneous or cumulative losses for the algorithms with respect to a large family of loss functions satisfying the constraint introduced by  ...  ACKNOWLEDGMENTS The author thanks Leonid Gurvits for helpful discussion and Jason Catlett for reading an earlier draft of this paper.  ... 
doi:10.1007/3-540-59119-2_170 fatcat:4546gem3grb4rlc4qvg57ypyw4

Page 2593 of Mathematical Reviews Vol. , Issue 97D [page]

1997 Mathematical Reviews  
Summary: “This paper introduces a new family of deterministic and stochastic on-line prediction algorithms which perform well with respect to general loss functions, and analyzes their behavior in terms  ...  The work pro- vides necessary theoretical support for applications and further study of the stochastic Tabu search strategy.” 974:68195 68T05S 62C99 62L99 Yamanishi, Kenji (1-NEC; Princeton, NJ) On-line  ... 

Modeling of Random Delays in Networked Control Systems

Yuan Ge, Qigong Chen, Ming Jiang, Yiqing Huang
2013 Journal of Control Science and Engineering  
In this paper, four major delay models are surveyed including constant delay model, mutually independent stochastic delay model, Markov chain model, and hidden Markov model.  ...  In order to compensate for random delays which may lead to performance degradation and instability of NCSs, it is necessary to establish the mathematical model of random delays before compensation.  ...  The PG algorithm classified off-line the network traffic into some separate patterns to derive a less conservative NCS, and the PI algorithm searched on-line for a pattern representing the recent network  ... 
doi:10.1155/2013/383415 fatcat:t55zr26aqzaoffndeh32vpmivq

Sequential Ski Rental Problem [article]

Anant Shah, Arun Rajkumar
2021 arXiv   pre-print
Under certain stochastic assumptions on the experts who predict the buy costs, we develop online algorithms and prove regret bounds for the same.  ...  We consider a variant of this problem which we call the 'sequential ski-rental' problem.  ...  ACKNOWLEDGMENTS Arun Rajkumar thanks Robert Bosch Center for Data Science and Artificial Intelligence, Indian Institute of Technology Madras for financial support.  ... 
arXiv:2104.06050v2 fatcat:3v2loddrsngprpgkmcyakk25za

Page 4357 of Mathematical Reviews Vol. , Issue 96g [page]

1996 Mathematical Reviews  
all the time spent in the process.” 96g:68113 68T05 62A99 Yamanishi, Kenji (1-NEC; Princeton, NJ) A loss bound model for on-line stochastic prediction algorithms.  ...  For this setting, we propose a weighted-average- type on-line stochastic prediction algorithm WA, which can be regarded as a hybrid of the Bayes algorithm and a sequential real- valued parameter estimation  ... 

Robust Model Predictive Control with One Free Control Move for NCSs with Data Missing

Jimin Yu, Xiaogang Gong
2016 International Journal of Computer Applications  
General Terms Networked system, predictive control Keywords Data missing, polytoic model, state feedback.  ...  Furthermore, the corresponding problems about recursive feasibility and stochastic stability are established by a set of linear matrix inequalities.  ...  Model predictive control(MPC) has a well-earned reputation in recent years, it provides on-line solutions to optimal feedback control problems, so model predictive control can be seen as a circular operation  ... 
doi:10.5120/ijca2016911825 fatcat:gec33adx4ndxfi625kvzzdbumm

Competitive On-Line Statistics

Volodya Vovk
2001 International Statistical Review  
In this approach, which we call "competitive on-line statistics", it is not assumed that data are generated by some stochastic mechanism; the bounds derived for the performance of competitive on-line statistical  ...  A radically new approach to statistical modelling, which combines mathematical techniques of Bayesian statistics with the philosophy of the theory of competitive on-line algorithms, has arisen over the  ...  a mistake in an earlier draft).  ... 
doi:10.2307/1403814 fatcat:w5amdxxdarhklo5ijzckhxrnui

Competitive On-line Statistics

Volodya Vovk
2001 International Statistical Review  
In this approach, which we call "competitive on-line statistics", it is not assumed that data are generated by some stochastic mechanism; the bounds derived for the performance of competitive on-line statistical  ...  A radically new approach to statistical modelling, which combines mathematical techniques of Bayesian statistics with the philosophy of the theory of competitive on-line algorithms, has arisen over the  ...  a mistake in an earlier draft).  ... 
doi:10.1111/j.1751-5823.2001.tb00457.x fatcat:px3lnybizbhuzjlnmzt4vkux7a

Stochastic Structured Prediction under Bandit Feedback [article]

Artem Sokolov and Julia Kreutzer and Christopher Lo and Stefan Riezler
2016 arXiv   pre-print
feedback in form of a task loss evaluation of the predicted structure.  ...  Stochastic structured prediction under bandit feedback follows a learning protocol where on each of a sequence of iterations, the learner receives an input, predicts an output structure, and receives partial  ...  This research was supported in part by the German research foundation (DFG), and in part by a research cooperation grant with the Amazon Development Center Germany.  ... 
arXiv:1606.00739v2 fatcat:khddswq23vcy7esrn2rcbb2tmi

An Online Stochastic Kernel Machine for Robust Signal Classification [article]

Raghu G. Raj
2019 arXiv   pre-print
space, which efficiently models the observed stationary process.  ...  We present a novel variation of online kernel machines in which we exploit a consensus based optimization mechanism to guide the evolution of decision functions drawn from a reproducing kernel Hilbert  ...  a sequence of samples from which predictions are made one at a time.  ... 
arXiv:1905.07686v2 fatcat:ctg7radssndhjhswpjnn2yutye

When Do Jumps Matter for Portfolio Optimization?

Marius Ascheberg, Nicole Branger, Holger Kraft
2013 Social Science Research Network  
Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Bounds on Relative Wealth Equivalent Loss When the investor relies on the approximating strategy from solving (6) , he suffers a utility loss.  ... 
doi:10.2139/ssrn.2257689 fatcat:juoawyhblfbb3exa2d537rsika

When do jumps matter for portfolio optimization?

Marius Ascheberg, Nicole Branger, Holger Kraft, Frank Thomas Seifried
2016 Quantitative finance (Print)  
Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Bounds on Relative Wealth Equivalent Loss When the investor relies on the approximating strategy from solving (6) , he suffers a utility loss.  ... 
doi:10.1080/14697688.2015.1131844 fatcat:l7j7rlhwznctvmkfno3p34e6dq

When Do Jumps Matter for Portfolio Optimization?

Marius Ascheberg, Nicole Branger, Holger Kraft
2013 Social Science Research Network  
Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Furthermore, we provide explicit bounds on the true optimal strategy and the relative wealth equivalent loss that do not rely on results from the true model.  ...  Bounds on Relative Wealth Equivalent Loss When the investor relies on the approximating strategy from solving (6) , he suffers a utility loss.  ... 
doi:10.2139/ssrn.2259630 fatcat:dyuxopsxbnbvtm36uiv6jhlxmq

Learning Fixation Point Strategy for Object Detection and Classification [article]

Jie Lyu, Zejian Yuan, Dapeng Chen
2017 arXiv   pre-print
Our method can predict a precise bounding box on an image, and achieve high speed on large images without pooling operations.  ...  On training, we present a hybrid loss function to learn the parameters of the multi-task network end-to-end.  ...  On training, we propose a multi-task loss function to optimize the network end-to-end, and combine stochastic and object-awareness (SA) strategy for learning fixation prediction.  ... 
arXiv:1712.06897v1 fatcat:pvknyqxdlnellhu6u4lxmib27m
« Previous Showing results 1 — 15 out of 74,101 results