Publications
49 results found
Flaxman S, Sejdinovic D, Cunningham JP, et al., 2016, Bayesian Learning of Kernel Embeddings, UAI'16
Kernel methods are one of the mainstays of machine learning, but the problemof kernel learning remains challenging, with only a few heuristics and verylittle theory. This is of particular importance in methods based on estimationof kernel mean embeddings of probability measures. For characteristic kernels,which include most commonly used ones, the kernel mean embedding uniquelydetermines its probability measure, so it can be used to design a powerfulstatistical testing framework, which includes nonparametric two-sample andindependence tests. In practice, however, the performance of these tests can bevery sensitive to the choice of kernel and its lengthscale parameters. Toaddress this central issue, we propose a new probabilistic model for kernelmean embeddings, the Bayesian Kernel Embedding model, combining a Gaussianprocess prior over the Reproducing Kernel Hilbert Space containing the meanembedding with a conjugate likelihood function, thus yielding a closed formposterior over the mean embedding. The posterior mean of our model is closelyrelated to recently proposed shrinkage estimators for kernel mean embeddings,while the posterior uncertainty is a new, interesting feature with variouspossible applications. Critically for the purposes of kernel learning, ourmodel gives a simple, closed form marginal pseudolikelihood of the observeddata given the kernel hyperparameters. This marginal pseudolikelihood caneither be optimized to inform the hyperparameter choice or fully Bayesianinference can be used.
Filippi S, Barnes CP, Kirk PDW, et al., 2016, Robustness of MEK-ERK Dynamics and Origins of Cell-to-Cell Variability in MAPK Signaling, CellReports
Mahon SSM, Lenive O, Filippi S, et al., 2015, Information processing by simple molecular motifs and susceptibility to noise, Journal of The Royal Society Interface
Bhatnagar N, Perkins K, Filippi S, et al., 2014, Clinical and Hematologic Impact of Fetal and Perinatal Variables on Mutant <i>GATA1</i> Clone Size in Neonates with Down Syndrome, BLOOD, Vol: 124, ISSN: 0006-4971
Mc Mahon SS, Sim A, Filippi S, et al., 2014, Information theory and signal transduction systems: From molecular information processing to network inference, SEMINARS IN CELL & DEVELOPMENTAL BIOLOGY, Vol: 35, Pages: 98-108, ISSN: 1084-9521
- Author Web Link
- Cite
- Citations: 34
MacLean AL, Filippi S, Stumpf MPH, 2014, The ecology in the hematopoietic stem cell niche determines the clinical outcome in chronic myeloid leukemia, PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, Vol: 111, Pages: 3883-3888, ISSN: 0027-8424
- Author Web Link
- Cite
- Citations: 33
Liepe J, Kirk P, Filippi S, et al., 2014, A framework for parameter estimation and model selection from experimental data in systems biology using approximate Bayesian computation, NATURE PROTOCOLS, Vol: 9, Pages: 439-456, ISSN: 1754-2189
- Author Web Link
- Cite
- Citations: 132
Silk D, Filippi S, Stumpf MPH, 2013, Optimizing threshold-schedules for sequential approximate Bayesian computation: applications to molecular systems, Statistical Applications in Genetics and Molecular Biology, Vol: 12, Pages: 603-618, ISSN: 2194-6302
The likelihood–free sequential Approximate Bayesian Computation (ABC) algorithms are increasingly popular inference tools for complex biological models. Such algorithms proceed by constructing a succession of probability distributions over the parameter space conditional upon the simulated data lying in an ε–ball around the observed data, for decreasing values of the threshold ε. While in theory, the distributions (starting from a suitably defined prior) will converge towards the unknown posterior as ε tends to zero, the exact sequence of thresholds can impact upon the computational efficiency and success of a particular application. In particular, we show here that the current preferred method of choosing thresholds as a pre-determined quantile of the distances between simulated and observed data from the previous population, can lead to the inferred posterior distribution being very different to the true posterior. Threshold selection thus remains an important challenge. Here we propose that the threshold–acceptance rate curve may be used to determine threshold schedules that avoid local optima, while balancing the need to minimise the threshold with computational efficiency. Furthermore, we provide an algorithm based upon the unscented transform, that enables the threshold–acceptance rate curve to be efficiently predicted in the case of deterministic and stochastic state space models.
Liepe J, Filippi S, Komorowski ML, et al., 2013, Maximizing the Information Content of Experiments in Systems Biology, PLoS computational biology
Filippi S, Barnes CP, Cornebise J, et al., 2013, On optimality of kernels for approximate Bayesian computation using sequential Monte Carlo, STATISTICAL APPLICATIONS IN GENETICS AND MOLECULAR BIOLOGY, Vol: 12, ISSN: 2194-6302
- Author Web Link
- Cite
- Citations: 50
Silk D, Filippi S, Stumpf MPH, 2013, Optimizing threshold-schedules for sequential approximate Bayesian computation: applications to molecular systems, STATISTICAL APPLICATIONS IN GENETICS AND MOLECULAR BIOLOGY, Vol: 12, Pages: 603-618, ISSN: 2194-6302
- Author Web Link
- Cite
- Citations: 32
Barnes CP, Filippi S, Stumpf MPH, et al., 2012, Considerate approaches to constructing summary statistics for ABC model selection, STATISTICS AND COMPUTING, Vol: 22, Pages: 1181-1197, ISSN: 0960-3174
- Author Web Link
- Cite
- Citations: 41
Roy A, Cowan G, Mead AJ, et al., 2012, Perturbation of fetal liver hematopoietic stem and progenitor cell development by trisomy 21, Proceedings of the National Academy of Sciences of the United States of America
Barnes C, Filippi S, Stumpf MPH, et al., 2012, Considerate approaches to achieving sufficiency for ABC model selection, Statistics and Computing, Vol: 22, Pages: 1181-1197, ISSN: 0960-3174
For nearly any challenging scientific problemevaluation of the likelihood is problematic if not impossible.Approximate Bayesian computation (ABC) allowsus to employ the whole Bayesian formalism to problemswhere we can use simulations from a model, but cannotevaluate the likelihood directly. When summary statistics ofreal and simulated data are compared—rather than the datadirectly—information is lost, unless the summary statisticsare sufficient. Sufficient statistics are, however, not commonbut without them statistical inference in ABC inferencesare to be considered with caution. Previously other authorshave attempted to combine different statistics in order toconstruct (approximately) sufficient statistics using searchand information heuristics. Here we employ an informationtheoreticalframework that can be used to construct appropriate(approximately sufficient) statistics by combining differentstatistics until the loss of information is minimized.We start from a potentially large number of different statisticsand choose the smallest set that captures (nearly) thesame information as the complete set. We then demonstratethat such sets of statistics can be constructed for both parameterestimation and model selection problems, and we applyour approach to a range of illustrative and real-world modelselection problems.
Roy A, Cowan G, Mead AJ, et al., 2012, Perturbation of fetal liver hematopoietic stem and progenitor cell development by trisomy 21., Proceedings of the National Academy of Sciences
Filippi S, Cappe O, Garivier A, 2011, Optimally Sensing a Single Channel Without Prior Information: The Tiling Algorithm and Regret Bounds, IEEE Journal of Selected Topics in Signal Processing
Filippi S, Cappe O, Garivier A, 2010, Optimism in Reinforcement Learning and Kullback-Leibler Divergence, ALLERTON 2010
We consider model-based reinforcement learning in finite Markov De- cisionProcesses (MDPs), focussing on so-called optimistic strategies. In MDPs,optimism can be implemented by carrying out extended value it- erations under aconstraint of consistency with the estimated model tran- sition probabilities.The UCRL2 algorithm by Auer, Jaksch and Ortner (2009), which follows thisstrategy, has recently been shown to guarantee near-optimal regret bounds. Inthis paper, we strongly argue in favor of using the Kullback-Leibler (KL)divergence for this purpose. By studying the linear maximization problem underKL constraints, we provide an ef- ficient algorithm, termed KL-UCRL, forsolving KL-optimistic extended value iteration. Using recent deviation boundson the KL divergence, we prove that KL-UCRL provides the same guarantees asUCRL2 in terms of regret. However, numerical experiments on classicalbenchmarks show a significantly improved behavior, particularly when the MDPhas reduced connectivity. To support this observation, we provide elements ofcom- parison between the two algorithms based on geometric considerations.
Filippi S, Cappe O, Garivier A, et al., 2010, Parametric bandits: The generalized linear case, Neural Information Processing Systems (NIPS’2010)
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.