6,464 Hits in 4.5 sec

Learning with attribute costs

Haim Kaplan, Eyal Kushilevitz, Yishay Mansour
2005 Proceedings of the thirty-seventh annual ACM symposium on Theory of computing - STOC '05  
We study an extension of the "standard" learning models to settings where observing the value of an attribute has an associated cost (which might be different for different attributes).  ...  The goal is to design a strategy to decide what attribute of x to observe next so as to minimize the expected evaluation cost of f (x).  ...  The most relevant work in the machine learning literature is the work of Madani et al. [14] which describes a fixed budget learning setting.  ... 
doi:10.1145/1060590.1060644 dblp:conf/stoc/KaplanKM05 fatcat:xrcrhq35offw5bwznuinrllhrq

Parallel bandit architecture based on laser chaos for reinforcement learning [article]

Takashi Urushibara, Nicolas Chauvet, Satoshi Kochi, Satoshi Sunada, Kazutaka Kanno, Atsushi Uchida, Ryoichi Horisaki, Makoto Naruse
2022 arXiv   pre-print
The architecture of Q-learning, however, does not fit well photonic implementations due to its separation of update rule and the action selection.  ...  However, reinforcement learning could involve a massive number of states, unlike previously demonstrated bandit problems where the number of states is only one.  ...  We use the Nelder-Mead method for the parameter optimization of Q-learning. We do not need the Golden section search because Q-learning does not have variations of random numbers.  ... 
arXiv:2205.09543v1 fatcat:mxnd3yg2arcy7lu5optjzhkkpa

Modeling Conceptual Understanding in Image Reference Games [article]

Rodolfo Corona, Stephan Alaniz, Zeynep Akata
2019 arXiv   pre-print
Our experiments on three benchmark image/attribute datasets suggest that our learner indeed encodes information directly pertaining to the understanding of other agents, and that leveraging this information  ...  An agent who interacts with a wide population of other agents needs to be aware that there may be variations in their understanding of the world.  ...  AI (XAI) does not explicitly reason about user understanding when generating explanations for model decisions.  ... 
arXiv:1910.04872v2 fatcat:brjwtw27sja7hb6vztaxc3s4t4

Anytime Learning of Decision Trees

Saher Esmeir, Shaul Markovitch
2007 Journal of machine learning research  
Even the few non-greedy learners cannot learn good trees when the concept is difficult.  ...  The majority of existing algorithms for learning decision trees are greedy-a tree is induced topdown, making locally optimal decisions at each node.  ...  Acknowledgments This work was partially supported by funding from the EC-sponsored MUSCLE Network of Excellence (FP6-507752).  ... 
dblp:journals/jmlr/EsmeirM07 fatcat:f345cdficbfnpc6f4r4ufy5n4i

Greed in Hamka's Novel Merantau ke Deli

Pardi Pardi
2021 International Journal of English and Applied Linguistics (IJEAL)  
The greediness of a beautiful woman also happens when he meets Mariatun, a beautiful young woman from the Minangkabaunese tribe.  ...  The greediness on poverty showed by Mariatun for taking the batik and also for dominating Leman's poverty.  ...  Leman felt for Mariatun when he came back to his village with Poniem. It is a tradition that when a Minangkabaunese man succeeds in merantau, he must come back to the village to show his success.  ... 
doi:10.47709/ijeal.v1i2.1209 fatcat:ipzgxk7id5ba5c4mkq2gmiswiy

Causal Models for Real Time Bidding with Repeated User Interactions

Martin Bompaire, Alexandre Gilotte, Benjamin Heymann
2021 Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining  
However, this greedy approach is too naive when several displays are shown to the same user.  ...  The purpose of the present paper is to discuss how such an estimation should be made when a user has already been shown one or more displays.  ...  The bidder can make a marketing campaign succeed or fail, because it decides for whom, when, where and at which price to buy a banner. * The three authors contributed equally to this research.  ... 
doi:10.1145/3447548.3467280 fatcat:ormqoi4q2rfvnavcpqlldw665e

An Encoder-Decoder Model for Visual Question Answering in the Medical Domain

Imane Allaouzi, Mohamed Ben Ahmed, Badr Benamrou
2019 Conference and Labs of the Evaluation Forum  
This paper describes our participation in the task of VQA-Med of ImageCLEF 2019.  ...  The answer generation is accomplished by the greedy search algorithm, which predicts the next word based on the previously generated words.  ...  This type of model often overfits when training on small datasets. To prevent overfitting, the best solution is to use the transfer learning technique.  ... 
dblp:conf/clef/AllaouziAB19 fatcat:wblweuof7zcwpdv4mg5fvvvv2a

A Bayesian Approach to Imitation in Reinforcement Learning

Bob Price, Craig Boutilier
2003 International Joint Conference on Artificial Intelligence  
In multiagent environments, forms of social learning such as teaching and imitation have been shown to aid the transfer of knowledge from experts to learners in reinforcement learning (RL).  ...  We recast the problem of imitation in a Bayesian framework.  ...  sober fact that the number of states is generally exponential in the number of the attributes defining a learning problem.  ... 
dblp:conf/ijcai/PriceB03 fatcat:ktycd7mwh5dghedskoqnkq6ziq

Configuration REcognition Model for Complex Reverse Engineering Methods: 2(CREM) [chapter]

Karim Hadjar, Oliver Hitz, Lyse Robadey, Rolf Ingold
2002 Lecture Notes in Computer Science  
The first prototype of 2(CREM) has been tested on four different phases of newspaper image analysis: line segment recognition, frame recognition, line merging into blocks, and logical labeling.  ...  This paper describes 2(CREM), a recognition method to be applied on documents with complex structures allowing incremental learning in an interactive environment.  ...  Some classification tasks use a subset of the available attributes. In 2(CREM), such attributes are called "relevant attributes".  ... 
doi:10.1007/3-540-45869-7_50 fatcat:p5njnafsvfcjrauosm2hryoxlq

Uncertainty and Reassurance in International Politics

Shiping Tang, Evan Braden Montgomery
2007 International Security  
In a recent article, Evan Montgomery addresses the question of how states cope with uncertainty about other states' intentions in international politics through reassurance. 1 He ªnds that because of a  ...  variety of constraining factors, attempts at reassurance have been rare and largely unsuccessful.  ...  For example, rather than call attention to the obvious dangers of exploitation, he simply notes that, when this occurs, the signaling state will learn that its adversary does indeed have malign motives  ... 
doi:10.1162/isec.2007.32.1.193 fatcat:c2elxp2rjrcxxjbiryunf6cvoa

Content Recommendation for Viral Social Influence

Sergei Ivanov, Konstantinos Theocharidis, Manolis Terrovitis, Panagiotis Karras
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
Greedy solution.  ...  In this paper, we address the natural problem that arises in such circumstances: Suggest content, expressed as a limited set of attributes, for a creative promotion campaign that starts out from a given  ...  , or even de ne a probability distribution function over the set of all attributes [7] , to be learned by historical logs.  ... 
doi:10.1145/3077136.3080788 dblp:conf/sigir/0002TTK17 fatcat:4cxecggekrctjk44zew6kpuzoq

Carousel Personalization in Music Streaming Apps with Contextual Bandits [article]

Walid Bendada and Guillaume Salha and Théo Bontempelli
2020 arXiv   pre-print
We empirically show the effectiveness of our framework at capturing characteristics of real-world carousels by addressing a large-scale playlist recommendation task on a global music streaming mobile app  ...  However, selecting the most relevant items (albums, artists, playlists...) to display in these carousels is a challenging task, as items are numerous and as users have different preferences.  ...  This result validates the relevance of capturing such phenomenon for our carousel-based personalization problem.  ... 
arXiv:2009.06546v1 fatcat:kosq7gqqgvdujo3fusww6lvngq

Induction of logic programs: FOIL and related systems

J. R. Quinlan, R. M. Cameron-Jones
1995 New generation computing  
foil is a rst-order learning system that uses information in a collection of relations to construct theories expressed in a dialect of Prolog.  ...  We present examples of tasks tackled by foil and of systems that adapt and extend its approach.  ...  We are grateful to William Cohen and Stephen Muggleton for most helpful comments on a draft of this paper.  ... 
doi:10.1007/bf03037228 fatcat:64xtevk4j5dezmsxjvk72hbapy

Learning Generative Deception Strategies in Combinatorial Masking Games [article]

Junlin Wu, Charles Kamhoua, Murat Kantarcioglu, Yevgeniy Vorobeychik
2022 arXiv   pre-print
We present a novel game-theoretic model of the resulting defender-attacker interaction, where the defender chooses a subset of attributes to mask, while the attacker responds by choosing an exploit to  ...  First, we show that the problem of computing an equilibrium of the resulting zero-sum defender-attacker game can be represented as a linear program with a combinatorial number of system configuration variables  ...  When there are multiple devices on the network, we represent each device by x k . However, we will omit this superscript when it is either not relevant, or not important.  ... 
arXiv:2109.11637v2 fatcat:v5nf4recuvde3kidxht3amsiki

The bias bias

Henry Brighton, Gerd Gigerenzer
2015 Journal of Business Research  
Here, we address the question of when and why simple models succeedor failby framing the forecasting problem in terms of the bias-variance dilemma.  ...  Using the study of cognitive heuristics, we discuss how to reduce variance by ignoring weights, attributes, and dependencies between attributes, and thus make better decisions.  ...  A number of studies have attempted to explain when and why takethe-best succeeds.  ... 
doi:10.1016/j.jbusres.2015.01.061 fatcat:6fsch64ekrf4xeloyvolbgbnqq
« Previous Showing results 1 — 15 out of 6,464 results