21,346 Hits in 4.8 sec

Guest editor's introduction

Lisa Hellerstein
1994 Machine Learning  
Schapire, and L. M. Sellie, introduces a generalization of the PAC model called the agnostic learning model.  ...  The extensive research generated by the model attests to its degree of success.  ...  The PAC learning model and its variants (including agnostic learning) model learning from random data.  ... 
doi:10.1007/bf00993467 fatcat:ww4ry7ctxfh7zd5wv7sg7n6y6a

PAC Bayesian Performance Guarantees for Deep (Stochastic) Networks in Medical Imaging [article]

Anthony Sicilia, Xingchen Zhao, Anastasia Sosnovskikh, Seong Jae Hwang
2021 arXiv   pre-print
Still, a "thorn in the side" of the deep learning movement is the argument that deep networks are prone to overfitting and are thus unable to generalize well when datasets are small (as is common in medical  ...  Application of deep neural networks to medical imaging tasks has in some sense become commonplace.  ...  This work is supported by the University of Pittsburgh Alzheimer Disease Research Center Grant (P30 AG066468).  ... 
arXiv:2104.05600v2 fatcat:nmpd7gdclnhf5hvmjrxbljnrga

Statistical Query Learning (1998; Kearns) [chapter]

Vitaly Feldman
2014 Encyclopedia of Algorithms  
In subsequent work the model of Kearns has been extended to other settings and found a number of additional applications in machine learning and theoretical computer science.  ...  Statistical query (SQ) learning is a natural restriction of PAC learning that models algorithms that use statistical properties of a data set rather than individual examples.  ...  Therefore the SQ model is a restriction of the PAC model.  ... 
doi:10.1007/978-3-642-27848-8_401-2 fatcat:wpv4onrvlnfiph4zr5wpaj5zvy

Statistical Query Learning [chapter]

Vitaly Feldman
2008 Encyclopedia of Algorithms  
In subsequent work the model of Kearns has been extended to other settings and found a number of additional applications in machine learning and theoretical computer science.  ...  Statistical query (SQ) learning is a natural restriction of PAC learning that models algorithms that use statistical properties of a data set rather than individual examples.  ...  Therefore the SQ model is a restriction of the PAC model.  ... 
doi:10.1007/978-0-387-30162-4_401 fatcat:cerelbnlzjffxniof5arf474d4

Page 5105 of Mathematical Reviews Vol. , Issue 93i [page]

1993 Mathematical Reviews  
Costica Cazacu (R-IASI) 93i:68149 68T05 92B20 Haussler, David (1-UCSC-CE) Decision-theoretic generalizations of the PAC model for neural net and other learning applications.  ...  (English summary) Inform. and Comput. 100 (1992), no. 1, 78-150. In the past few years, a fruitful theory of learning has developed, based on the so-called PAC model [L. G. Valiant, Comm.  ... 

Integrating AI into Radiology workflow: Levels of research, production, and feedback maturity [article]

Engin Dikici, Matthew Bigelow, Luciano M. Prevedello, Richard D. White, Barbaros Selnur Erdal
2019 arXiv   pre-print
A case study (i.e., detection of brain metastases with T1-weighted contrast-enhanced 3D MRI) illustrates the deployment details of a particular AI-based application according to the aforementioned maturity  ...  It is shown that the given AI application significantly improves with the feedback coming from radiologists; the number of incorrectly detected brain metastases (false positives) reduces from 14.2 to 9.12  ...  After the result dataset is generated, it is sent to a separate Research-PACS; this enables the official EMR in PACS to remain intact.  ... 
arXiv:1910.06424v1 fatcat:6ahoragrtza67ijof2sfvbffse

Identifying Offline Muscle Strength Profiles Sufficient for Short-Duration Fes-Lce Exercise: A Pac Learning Model Approach

Randy D. Trumbower, Sanguthevar Rajasekaran, Pouran D. Faghri
2006 Journal of clinical monitoring and computing  
To develop and test a Probably Approximately Correct (PAC) learning model as a predictor of thigh muscle strengths sufficient for short-duration FES-LCE exercise and compare the model's performance with  ...  PAC and ANN learning models correctly identified 100% of the training examples. PAC's average performance on the validation set was 93.1%.  ...  However, it is not clear if a PAC learning model is feasible for classifying muscle strengths, because it has not been previously assessed for this type of clinical application.  ... 
doi:10.1007/s10877-006-9023-2 pmid:16775658 fatcat:y625ixne3rgofhzgxwf56sk2t4

Learning under Model Misspecification: Applications to Variational and Ensemble methods [article]

Andres R. Masegosa
2020 arXiv   pre-print
In this work, we present a novel analysis of the generalization performance of Bayesian model averaging under model misspecification and i.i.d. data using a new family of second-order PAC-Bayes bounds.  ...  Using novel second-order PAC-Bayes bounds, we derive a new family of Bayesian-like algorithms, which can be implemented as variational and ensemble methods.  ...  But it was in [18] were a neat connection was established between Bayesian learning and PAC-Bayesian theory.  ... 
arXiv:1912.08335v5 fatcat:oci35czdhzb7fnagrn5r2gzmqy

Evolvability need not imply learnability [article]

Nisheeth Srivastava
2009 arXiv   pre-print
The implications of the latter case on the prospects of learning in complex hypothesis spaces is briefly examined.  ...  We further show that this result is insufficient to prove the PAC-learnability of monotone Boolean functions, thereby demonstrating a counter-example to a recent claim to the contrary.  ...  Informally, it may be seen that finding a DNF that while learning theory and its applications seek to model and learn a DNF that is structurally similar to the target function in Nature, genotypes and  ... 
arXiv:0904.0648v1 fatcat:qbknbnzlsjhpfjz7gh47wmma5q

Natural Brain-Inspired Intelligence for Non-Gaussian and Nonlinear Environments with Finite Memory

Naghshvarianjahromi, Kumar, Deen
2020 Applied Sciences  
Therefore, it can easily upgrade the conventional systems to a smart one for autonomic CDM applications.  ...  The simple low-complexity algorithmic design of the proposed CDM system can make it suitable for real-time applications.  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/app10031150 fatcat:ncafexnnsbb7pp6kqnvqktvcrq

Hardness of Proper Learning (1988; Pitt, Valiant) [chapter]

Vitaly Feldman
2014 Encyclopedia of Algorithms  
The work of Pitt and Valiant [16] deals with learning Boolean functions in the Probably Approximately Correct (PAC) learning model introduced by Valiant [17].  ...  A learning algorithm in Valiant's original model is given random examples of a function f : {0, 1} n → {0, 1} from a representation class F and produces a hypothesis h ∈ F that closely approximates f .  ...  OPEN PROBLEMS A number of problems related to proper learning in the PAC model and its extensions are open.  ... 
doi:10.1007/978-3-642-27848-8_177-2 fatcat:fo6epnlbqndvzjlv6uqiojmflq


2005 Biocomputing 2006  
Bayesian networks (BNs) and dynamic Bayesian networks (DBNs) are becoming more widely used as a way to learn various types of networks, including cellular signaling networks 16,14 , from high-throughput  ...  We present a theoretical analysis on the ability of DBNs without hidden variables to learn from proteomic time series data.  ...  The models defined in this paper are a natural application of existing PAC models to DBN learning.  ... 
doi:10.1142/9789812701626_0025 fatcat:iprb5oq5kfhp5pm4x7vg7gico4

Page 7822 of Mathematical Reviews Vol. , Issue 97M [page]

1997 Mathematical Reviews  
Summary: “In this paper we study a new view of the PAC-learning model in which the examples are more complicated than in the standard model.  ...  It gives a detailed presentation of many small examples and a brief overview of some industrial applications. It introduces the formal analysis methods.  ... 

Assessing Test Adequacy for Black-Box Systems without Specifications [chapter]

Neil Walkinshaw
2011 Lecture Notes in Computer Science  
Several researchers have noted that there is a duality between these testing problems and the problem of inductive inference (learning a model of a hidden system from a given set of examples).  ...  It is impossible to tell how many examples will be required to infer an accurate model, and there is no basis for telling how complete a given set of examples is.  ...  Acknowledgements Much of the background material that relates inductive inference to software testing was influenced by discussions with Gordon Fraser at Saarland University in Saarbrücken.  ... 
doi:10.1007/978-3-642-24580-0_15 fatcat:wwqbvfxw3ngvnm4uihrk5zyyly

Probably Approximately Correct, A Book Review

Marcus Feldman
2014 Notices of the American Mathematical Society  
Chapters 6, 7, and 8 delve into the potential applications of PAC learning to biological evolution, human reasoning and cognition, and human emotion and language.  ...  (a) understand it and (b) see why it might have greater utility than the mainstream mathematical modeling in their fields.  ... 
doi:10.1090/noti1177 fatcat:mns374c7wbecdftkv6lpfan4vm
« Previous Showing results 1 — 15 out of 21,346 results