Showing 1–43 of 43 results for author: Franks, J

Towards Graph Foundation Models: A Study on the Generalization of Positional and Structural Encodings

Authors: Billy Joe Franks, Moshe Eliasof, Semih Cantürk, Guy Wolf, Carola-Bibiane Schönlieb, Sophie Fellenz, Marius Kloft

Abstract: Recent advances in integrating positional and structural encodings (PSEs) into graph neural networks (GNNs) have significantly enhanced their performance across various graph learning tasks. However, the general applicability of these encodings and their potential to serve as foundational representations for graphs remain uncertain. This paper investigates the fine-tuning efficiency, scalability w… ▽ More Recent advances in integrating positional and structural encodings (PSEs) into graph neural networks (GNNs) have significantly enhanced their performance across various graph learning tasks. However, the general applicability of these encodings and their potential to serve as foundational representations for graphs remain uncertain. This paper investigates the fine-tuning efficiency, scalability with sample size, and generalization capability of learnable PSEs across diverse graph datasets. Specifically, we evaluate their potential as universal pre-trained models that can be easily adapted to new tasks with minimal fine-tuning and limited data. Furthermore, we assess the expressivity of the learned representations, particularly, when used to augment downstream GNNs. We demonstrate through extensive benchmarking and empirical analysis that PSEs generally enhance downstream models. However, some datasets may require specific PSE-augmentations to achieve optimal performance. Nevertheless, our findings highlight their significant potential to become integral components of future graph foundation models. We provide new insights into the strengths and limitations of PSEs, contributing to the broader discourse on foundation models in graph learning. △ Less

Submitted 3 March, 2025; v1 submitted 10 December, 2024; originally announced December 2024.

Comments: Published at TMLR (https://openreview.net/forum?id=mSoDRZXsqj)

Weisfeiler-Leman at the margin: When more expressivity matters

Authors: Billy J. Franks, Christopher Morris, Ameya Velingker, Floris Geerts

Abstract: The Weisfeiler-Leman algorithm ($1$-WL) is a well-studied heuristic for the graph isomorphism problem. Recently, the algorithm has played a prominent role in understanding the expressive power of message-passing graph neural networks (MPNNs) and being effective as a graph kernel. Despite its success, $1$-WL faces challenges in distinguishing non-isomorphic graphs, leading to the development of mor… ▽ More The Weisfeiler-Leman algorithm ($1$-WL) is a well-studied heuristic for the graph isomorphism problem. Recently, the algorithm has played a prominent role in understanding the expressive power of message-passing graph neural networks (MPNNs) and being effective as a graph kernel. Despite its success, $1$-WL faces challenges in distinguishing non-isomorphic graphs, leading to the development of more expressive MPNN and kernel architectures. However, the relationship between enhanced expressivity and improved generalization performance remains unclear. Here, we show that an architecture's expressivity offers limited insights into its generalization performance when viewed through graph isomorphism. Moreover, we focus on augmenting $1$-WL and MPNNs with subgraph information and employ classical margin theory to investigate the conditions under which an architecture's increased expressivity aligns with improved generalization performance. In addition, we show that gradient flow pushes the MPNN's weights toward the maximum margin solution. Further, we introduce variations of expressive $1$-WL-based kernel and MPNN architectures with provable generalization properties. Our empirical study confirms the validity of our theoretical findings. △ Less

Submitted 28 May, 2024; v1 submitted 12 February, 2024; originally announced February 2024.

Comments: Accepted at ICML 2024. arXiv admin note: text overlap with arXiv:2301.11039

Deep Anomaly Detection on Tennessee Eastman Process Data

Authors: Fabian Hartung, Billy Joe Franks, Tobias Michels, Dennis Wagner, Philipp Liznerski, Steffen Reithermann, Sophie Fellenz, Fabian Jirasek, Maja Rudolph, Daniel Neider, Heike Leitte, Chen Song, Benjamin Kloepper, Stephan Mandt, Michael Bortz, Jakob Burger, Hans Hasse, Marius Kloft

Abstract: This paper provides the first comprehensive evaluation and analysis of modern (deep-learning) unsupervised anomaly detection methods for chemical process data. We focus on the Tennessee Eastman process dataset, which has been a standard litmus test to benchmark anomaly detection methods for nearly three decades. Our extensive study will facilitate choosing appropriate anomaly detection methods in… ▽ More This paper provides the first comprehensive evaluation and analysis of modern (deep-learning) unsupervised anomaly detection methods for chemical process data. We focus on the Tennessee Eastman process dataset, which has been a standard litmus test to benchmark anomaly detection methods for nearly three decades. Our extensive study will facilitate choosing appropriate anomaly detection methods in industrial applications. △ Less

Submitted 10 March, 2023; originally announced March 2023.

Counting Parking Sequences and Parking Assortments Through Permutations

Authors: Spencer J. Franks, Pamela E. Harris, Kimberly Harry, Jan Kretschmann, Megan Vance

Abstract: Parking sequences (a generalization of parking functions) are defined by specifying car lengths and requiring that a car attempts to park in the first available spot after its preference. If it does not fit there, then a collision occurs and the car fails to park. In contrast, parking assortments generalize parking sequences (and parking functions) by allowing cars (also of assorted lengths) to se… ▽ More Parking sequences (a generalization of parking functions) are defined by specifying car lengths and requiring that a car attempts to park in the first available spot after its preference. If it does not fit there, then a collision occurs and the car fails to park. In contrast, parking assortments generalize parking sequences (and parking functions) by allowing cars (also of assorted lengths) to seek forward from their preference to identify a set of contiguous unoccupied spots in which they fit. We consider both parking sequences and parking assortments and establish that the number of preferences resulting in a fixed parking order $σ$ is related to the lengths of cars indexed by certain subsequences in $σ$. The sum of these numbers over all parking orders (i.e. permutations of $[n]$) yields new formulas for the total number of parking sequences and of parking assortments. △ Less

Submitted 25 January, 2023; originally announced January 2023.

Comments: 9 pages, 2 figures, 1 table

MSC Class: 05A15; 05A19

Ordinal Regression for Difficulty Estimation of StepMania Levels

Authors: Billy Joe Franks, Benjamin Dinkelmann, Sophie Fellenz, Marius Kloft

Abstract: StepMania is a popular open-source clone of a rhythm-based video game. As is common in popular games, there is a large number of community-designed levels. It is often difficult for players and level authors to determine the difficulty level of such community contributions. In this work, we formalize and analyze the difficulty prediction task on StepMania levels as an ordinal regression (OR) task.… ▽ More StepMania is a popular open-source clone of a rhythm-based video game. As is common in popular games, there is a large number of community-designed levels. It is often difficult for players and level authors to determine the difficulty level of such community contributions. In this work, we formalize and analyze the difficulty prediction task on StepMania levels as an ordinal regression (OR) task. We standardize a more extensive and diverse selection of this data resulting in five data sets, two of which are extensions of previous work. We evaluate many competitive OR and non-OR models, demonstrating that neural network-based models significantly outperform the state of the art and that StepMania-level data makes for an excellent test bed for deep OR models. We conclude with a user experiment showing our trained models' superiority over human labeling. △ Less

Submitted 23 January, 2023; originally announced January 2023.

Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images

Authors: Philipp Liznerski, Lukas Ruff, Robert A. Vandermeulen, Billy Joe Franks, Klaus-Robert Müller, Marius Kloft

Abstract: Due to the intractability of characterizing everything that looks unlike the normal data, anomaly detection (AD) is traditionally treated as an unsupervised problem utilizing only normal samples. However, it has recently been found that unsupervised image AD can be drastically improved through the utilization of huge corpora of random images to represent anomalousness; a technique which is known a… ▽ More Due to the intractability of characterizing everything that looks unlike the normal data, anomaly detection (AD) is traditionally treated as an unsupervised problem utilizing only normal samples. However, it has recently been found that unsupervised image AD can be drastically improved through the utilization of huge corpora of random images to represent anomalousness; a technique which is known as Outlier Exposure. In this paper we show that specialized AD learning methods seem unnecessary for state-of-the-art performance, and furthermore one can achieve strong performance with just a small collection of Outlier Exposure data, contradicting common assumptions in the field of AD. We find that standard classifiers and semi-supervised one-class methods trained to discern between normal samples and relatively few random natural images are able to outperform the current state of the art on an established AD benchmark with ImageNet. Further experiments reveal that even one well-chosen outlier sample is sufficient to achieve decent performance on this benchmark (79.3% AUC). We investigate this phenomenon and find that one-class methods are more robust to the choice of training outliers, indicating that there are scenarios where these are still more useful than standard classifiers. Additionally, we include experiments that delineate the scenarios where our results hold. Lastly, no training samples are necessary when one uses the representations learned by CLIP, a recent foundation model, which achieves state-of-the-art AD results on CIFAR-10 and ImageNet in a zero-shot setting. △ Less

Submitted 14 November, 2022; v1 submitted 23 May, 2022; originally announced May 2022.

Comments: 47 pages; extended experiments; published in Transactions on Machine Learning Research. arXiv admin note: substantial text overlap with arXiv:2006.00339

A systematic approach to random data augmentation on graph neural networks

Authors: Billy Joe Franks, Markus Anders, Marius Kloft, Pascal Schweitzer

Abstract: Random data augmentations (RDAs) are state of the art regarding practical graph neural networks that are provably universal. There is great diversity regarding terminology, methodology, benchmarks, and evaluation metrics used among existing RDAs. Not only does this make it increasingly difficult for practitioners to decide which technique to apply to a given problem, but it also stands in the way… ▽ More Random data augmentations (RDAs) are state of the art regarding practical graph neural networks that are provably universal. There is great diversity regarding terminology, methodology, benchmarks, and evaluation metrics used among existing RDAs. Not only does this make it increasingly difficult for practitioners to decide which technique to apply to a given problem, but it also stands in the way of systematic improvements. We propose a new comprehensive framework that captures all previous RDA techniques. On the theoretical side, among other results, we formally prove that under natural conditions all instantiations of our framework are universal. On the practical side, we develop a method to systematically and automatically train RDAs. This in turn enables us to impartially and objectively compare all existing RDAs. New RDAs naturally emerge from our approach, and our experiments demonstrate that they improve the state of the art. △ Less

Submitted 21 March, 2022; v1 submitted 8 December, 2021; originally announced December 2021.

Explainable Deep One-Class Classification

Authors: Philipp Liznerski, Lukas Ruff, Robert A. Vandermeulen, Billy Joe Franks, Marius Kloft, Klaus-Robert Müller

Abstract: Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away. Because this transformation is highly non-linear, finding interpretations poses a significant challenge. In this paper we present an explainable deep one-class classification method, Fully Convolutional Data Description (FCDD), where t… ▽ More Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away. Because this transformation is highly non-linear, finding interpretations poses a significant challenge. In this paper we present an explainable deep one-class classification method, Fully Convolutional Data Description (FCDD), where the mapped samples are themselves also an explanation heatmap. FCDD yields competitive detection performance and provides reasonable explanations on common anomaly detection benchmarks with CIFAR-10 and ImageNet. On MVTec-AD, a recent manufacturing dataset offering ground-truth anomaly maps, FCDD sets a new state of the art in the unsupervised setting. Our method can incorporate ground-truth anomaly maps during training and using even a few of these (~5) improves performance significantly. Finally, using FCDD's explanations we demonstrate the vulnerability of deep one-class classification models to spurious image features such as image watermarks. △ Less

Submitted 18 March, 2021; v1 submitted 3 July, 2020; originally announced July 2020.

Comments: 25 pages, published as a conference paper at ICLR 2021

Rethinking Assumptions in Deep Anomaly Detection

Authors: Lukas Ruff, Robert A. Vandermeulen, Billy Joe Franks, Klaus-Robert Müller, Marius Kloft

Abstract: Though anomaly detection (AD) can be viewed as a classification problem (nominal vs. anomalous) it is usually treated in an unsupervised manner since one typically does not have access to, or it is infeasible to utilize, a dataset that sufficiently characterizes what it means to be "anomalous." In this paper we present results demonstrating that this intuition surprisingly seems not to extend to d… ▽ More Though anomaly detection (AD) can be viewed as a classification problem (nominal vs. anomalous) it is usually treated in an unsupervised manner since one typically does not have access to, or it is infeasible to utilize, a dataset that sufficiently characterizes what it means to be "anomalous." In this paper we present results demonstrating that this intuition surprisingly seems not to extend to deep AD on images. For a recent AD benchmark on ImageNet, classifiers trained to discern between normal samples and just a few (64) random natural images are able to outperform the current state of the art in deep AD. Experimentally we discover that the multiscale structure of image data makes example anomalies exceptionally informative. △ Less

Submitted 27 January, 2023; v1 submitted 30 May, 2020; originally announced June 2020.

Comments: 17 pages; accepted at the ICML 2021 Workshop on Uncertainty & Robustness in Deep Learning; An extended Journal paper of this work has been published in Transactions on Machine Learning Research: arXiv:2205.11474

Markov chain Monte Carlo importance samplers for Bayesian models with intractable likelihoods

Authors: Jordan Franks

Abstract: We consider the efficient use of an approximation within Markov chain Monte Carlo (MCMC), with subsequent importance sampling (IS) correction of the Markov chain inexact output, leading to asymptotically exact inference. We detail convergence and central limit theorems for the resulting MCMC-IS estimators. We also consider the case where the approximate Markov chain is pseudo-marginal, requiring u… ▽ More We consider the efficient use of an approximation within Markov chain Monte Carlo (MCMC), with subsequent importance sampling (IS) correction of the Markov chain inexact output, leading to asymptotically exact inference. We detail convergence and central limit theorems for the resulting MCMC-IS estimators. We also consider the case where the approximate Markov chain is pseudo-marginal, requiring unbiased estimators for its approximate marginal target. Convergence results with asymptotic variance formulae are shown for this case, and for the case where the IS weights based on unbiased estimators are only calculated for distinct output samples of the so-called `jump' chain, which, with a suitable reweighting, allows for improved efficiency. As the IS type weights may assume negative values, extended classes of unbiased estimators may be used for the IS type correction, such as those obtained from randomised multilevel Monte Carlo. Using Euler approximations and coupling of particle filters, we apply the resulting estimator using randomised weights to the problem of parameter inference for partially observed Itô diffusions. Convergence of the estimator is verified to hold under regularity assumptions which do not require that the diffusion can be simulated exactly. In the context of approximate Bayesian computation (ABC), we suggest an adaptive MCMC approach to deal with the selection of a suitably large tolerance, with IS correction possible to finer tolerance, and with provided approximate confidence intervals. A prominent question is the efficiency of MCMC-IS compared to standard direct MCMC, such as pseudo-marginal, delayed acceptance, and ABC-MCMC. We provide a comparison criterion which generalises the covariance ordering to the IS setting. △ Less

Submitted 11 April, 2019; originally announced April 2019.

Comments: 41 pages. Introductory part of doctoral dissertation. Involves previous joint works with Matti Vihola, Jouni Helske, Ajay Jasra, and Kody J.H. Law. PhD thesis, University of Jyväskylä, 2019. Available online at http://urn.fi/URN:ISBN:978-951-39-7738-2

Report number: JYU Dissertations 79 MSC Class: 65C05

On the use of approximate Bayesian computation Markov chain Monte Carlo with inflated tolerance and post-correction

Authors: Matti Vihola, Jordan Franks

Abstract: Approximate Bayesian computation allows for inference of complicated probabilistic models with intractable likelihoods using model simulations. The Markov chain Monte Carlo implementation of approximate Bayesian computation is often sensitive to the tolerance parameter: low tolerance leads to poor mixing and large tolerance entails excess bias. We consider an approach using a relatively large tole… ▽ More Approximate Bayesian computation allows for inference of complicated probabilistic models with intractable likelihoods using model simulations. The Markov chain Monte Carlo implementation of approximate Bayesian computation is often sensitive to the tolerance parameter: low tolerance leads to poor mixing and large tolerance entails excess bias. We consider an approach using a relatively large tolerance for the Markov chain Monte Carlo sampler to ensure its sufficient mixing, and post-processing the output leading to estimators for a range of finer tolerances. We introduce an approximate confidence interval for the related post-corrected estimators, and propose an adaptive approximate Bayesian computation Markov chain Monte Carlo, which finds a `balanced' tolerance level automatically, based on acceptance rate optimisation. Our experiments show that post-processing based estimators can perform better than direct Markov chain targetting a fine tolerance, that our confidence intervals are reliable, and that our adaptive algorithm leads to reliable inference with little user specification. △ Less

Submitted 16 May, 2019; v1 submitted 1 February, 2019; originally announced February 2019.

Comments: 27 pages, 8 figures, 11 tables

Polygonal ${\mathbb Z}^2$-subshifts

Authors: John Franks, Bryna Kra

Abstract: Let ${\mathcal P}\subset{\mathbb Z}^2$ be a convex polygon with each vertex in it labeled by an element from a finite set and such that the labeling of each vertex $v\in {\mathcal P}$ is uniquely determined by the labeling of all other points in the polygon. We introduce a class of ${\mathbb Z}^2$-shift systems, the {\em polygonal shifts}, determined by such a polygon: these are shift systems such… ▽ More Let ${\mathcal P}\subset{\mathbb Z}^2$ be a convex polygon with each vertex in it labeled by an element from a finite set and such that the labeling of each vertex $v\in {\mathcal P}$ is uniquely determined by the labeling of all other points in the polygon. We introduce a class of ${\mathbb Z}^2$-shift systems, the {\em polygonal shifts}, determined by such a polygon: these are shift systems such that the restriction of any $x\in X$ to some polygon ${\mathcal P}$ has this property. These polygonal systems are related to various well studied classes of shift systems, including subshifts of finite type and algebraic shifts, but include many other systems. We give necessary conditions for a ${\mathbb Z}^2$-system $X$ to be polygonal, in terms of the nonexpansive subspaces of $X$, and under further conditions can give a complete characterization for such systems. △ Less

Submitted 30 October, 2019; v1 submitted 29 January, 2019; originally announced January 2019.

Unbiased inference for discretely observed hidden Markov model diffusions

Authors: Neil K. Chada, Jordan Franks, Ajay Jasra, Kody J. H. Law, Matti Vihola

Abstract: We develop a Bayesian inference method for diffusions observed discretely and with noise, which is free of discretisation bias. Unlike existing unbiased inference methods, our method does not rely on exact simulation techniques. Instead, our method uses standard time-discretised approximations of diffusions, such as the Euler--Maruyama scheme. Our approach is based on particle marginal Metropolis-… ▽ More We develop a Bayesian inference method for diffusions observed discretely and with noise, which is free of discretisation bias. Unlike existing unbiased inference methods, our method does not rely on exact simulation techniques. Instead, our method uses standard time-discretised approximations of diffusions, such as the Euler--Maruyama scheme. Our approach is based on particle marginal Metropolis--Hastings, a particle filter, randomised multilevel Monte Carlo, and importance sampling type correction of approximate Markov chain Monte Carlo. The resulting estimator leads to inference without a bias from the time-discretisation as the number of Markov chain iterations increases. We give convergence results and recommend allocations for algorithm inputs. Our method admits a straightforward parallelisation, and can be computationally efficient. The user-friendly approach is illustrated on three examples, where the underlying diffusion is an Ornstein--Uhlenbeck process, a geometric Brownian motion, and a 2d non-reversible Langevin equation. △ Less

Submitted 9 March, 2021; v1 submitted 26 July, 2018; originally announced July 2018.

Comments: 33 pages, 5 figures

MSC Class: 65C05 (primary); 60H35; 65C35; 65C40 (secondary)

Importance sampling correction versus standard averages of reversible MCMCs in terms of the asymptotic variance

Authors: Jordan Franks, Matti Vihola

Abstract: We establish an ordering criterion for the asymptotic variances of two consistent Markov chain Monte Carlo (MCMC) estimators: an importance sampling (IS) estimator, based on an approximate reversible chain and subsequent IS weighting, and a standard MCMC estimator, based on an exact reversible chain. Essentially, we relax the criterion of the Peskun type covariance ordering by considering two diff… ▽ More We establish an ordering criterion for the asymptotic variances of two consistent Markov chain Monte Carlo (MCMC) estimators: an importance sampling (IS) estimator, based on an approximate reversible chain and subsequent IS weighting, and a standard MCMC estimator, based on an exact reversible chain. Essentially, we relax the criterion of the Peskun type covariance ordering by considering two different invariant probabilities, and obtain, in place of a strict ordering of asymptotic variances, a bound of the asymptotic variance of IS by that of the direct MCMC. Simple examples show that IS can have arbitrarily better or worse asymptotic variance than Metropolis-Hastings and delayed-acceptance (DA) MCMC. Our ordering implies that IS is guaranteed to be competitive up to a factor depending on the supremum of the (marginal) IS weight. We elaborate upon the criterion in case of unbiased estimators as part of an auxiliary variable framework. We show how the criterion implies asymptotic variance guarantees for IS in terms of pseudo-marginal (PM) and DA corrections, essentially if the ratio of exact and approximate likelihoods is bounded. We also show that convergence of the IS chain can be less affected by unbounded high-variance unbiased estimators than PM and DA chains. △ Less

Submitted 24 March, 2020; v1 submitted 29 June, 2017; originally announced June 2017.

Comments: 27 pages

MSC Class: 60J22; 65C05

Journal ref: Stochastic Processes and their Applications, 2020

Notes on Chain Recurrence and Lyapunonv Functions

Authors: John Franks

Abstract: This short expository note provides an introduction to the concept of chain recurrence in topological dynamics and a proof of the existence complete Lyapunov functions for homeomorphisms of compact metric spaces due to Charles Conley. I have used it as supplementary material in introductory dynamics courses. This short expository note provides an introduction to the concept of chain recurrence in topological dynamics and a proof of the existence complete Lyapunov functions for homeomorphisms of compact metric spaces due to Charles Conley. I have used it as supplementary material in introductory dynamics courses. △ Less

Submitted 24 April, 2017; originally announced April 2017.

MSC Class: 37B25

Distortion and the automorphism group of a shift

Authors: Van Cyr, John Franks, Bryna Kra, Samuel Petite

Abstract: The set of automorphisms of a one-dimensional \shift $(X, σ)$ forms a countable, but often very complicated, group. For zero entropy shifts, it has recently been shown that the automorphism group is more tame. We provide the first examples of countable groups that cannot embed into the automorphism group of any zero entropy \shiftno. In particular, we show that the Baumslag-Solitar groups… ▽ More The set of automorphisms of a one-dimensional \shift $(X, σ)$ forms a countable, but often very complicated, group. For zero entropy shifts, it has recently been shown that the automorphism group is more tame. We provide the first examples of countable groups that cannot embed into the automorphism group of any zero entropy \shiftno. In particular, we show that the Baumslag-Solitar groups ${\rm BS}(1,n)$ and all other groups that contain exponentially distorted elements cannot embed into ${\rm Aut}(X)$ when $h_{\rm top}(X)=0$. We further show that distortion in nilpotent groups gives a nontrivial obstruction to embedding such a group in any low complexity shift. △ Less

Submitted 9 August, 2017; v1 submitted 17 November, 2016; originally announced November 2016.

The spacetime of a shift endomorphism

Authors: Van Cyr, John Franks, Bryna Kra

Abstract: The automorphism group of a one dimensional shift space over a finite alphabet exhibits different types of behavior: for a large class with positive entropy, it contains a rich collection of subgroups, while for many shifts of zero entropy, there are strong constraints on the automorphism group. We view this from a different perspective, considering a single automorphism (and sometimes endomorphis… ▽ More The automorphism group of a one dimensional shift space over a finite alphabet exhibits different types of behavior: for a large class with positive entropy, it contains a rich collection of subgroups, while for many shifts of zero entropy, there are strong constraints on the automorphism group. We view this from a different perspective, considering a single automorphism (and sometimes endomorphism) and studying the naturally associated two dimensional shift system. In particular, we describe the relation between nonexpansive subspaces in this two dimensional system and dynamical properties of an automorphism of the shift. △ Less

Submitted 9 August, 2017; v1 submitted 25 October, 2016; originally announced October 2016.

Importance sampling type estimators based on approximate marginal MCMC

Authors: Matti Vihola, Jouni Helske, Jordan Franks

Abstract: We consider importance sampling (IS) type weighted estimators based on Markov chain Monte Carlo (MCMC) targeting an approximate marginal of the target distribution. In the context of Bayesian latent variable models, the MCMC typically operates on the hyperparameters, and the subsequent weighting may be based on IS or sequential Monte Carlo (SMC), but allows for multilevel techniques as well. The I… ▽ More We consider importance sampling (IS) type weighted estimators based on Markov chain Monte Carlo (MCMC) targeting an approximate marginal of the target distribution. In the context of Bayesian latent variable models, the MCMC typically operates on the hyperparameters, and the subsequent weighting may be based on IS or sequential Monte Carlo (SMC), but allows for multilevel techniques as well. The IS approach provides a natural alternative to delayed acceptance (DA) pseudo-marginal/particle MCMC, and has many advantages over DA, including a straightforward parallelisation and additional flexibility in MCMC implementation. We detail minimal conditions which ensure strong consistency of the suggested estimators, and provide central limit theorems with expressions for asymptotic variances. We demonstrate how our method can make use of SMC in the state space models context, using Laplace approximations and time-discretised diffusions. Our experimental results are promising and show that the IS type approach can provide substantial gains relative to an analogous DA scheme, and is often competitive even without parallelisation. △ Less

Submitted 9 March, 2020; v1 submitted 8 September, 2016; originally announced September 2016.

Comments: 34 pages, 1 figure

Journal ref: Scand J Statist. 2020; 47: 1339-1376

Zero entropy subgroups of mapping class groups

Authors: John Franks, Kamlesh Parwani

Abstract: Let $M$ be a compact surface with boundary. We are interested in the question of how a group action on $M$ permutes a finite invariant set $X \subset int(M)$. More precisely, how the algebraic properties of the induced group of permutations of a finite invariant set affects the dynamical properties of the group. Our main result shows that in many circumstances if the induced permutation group is n… ▽ More Let $M$ be a compact surface with boundary. We are interested in the question of how a group action on $M$ permutes a finite invariant set $X \subset int(M)$. More precisely, how the algebraic properties of the induced group of permutations of a finite invariant set affects the dynamical properties of the group. Our main result shows that in many circumstances if the induced permutation group is not solvable then among the homeomorphisms in the group there must be one with a pseudo-Anosov component. We formulate this in terms of the mapping class group relative to the finite set and show the stronger result that in many circumstances (e.g. if $\partial M \ne \emptyset$) this mapping class group is itself solvable if it has no elements with pseudo-Anosov components. △ Less

Submitted 8 May, 2016; v1 submitted 8 February, 2015; originally announced February 2015.

Rotation Numbers for $S^2$ diffeomorphisms

Authors: John Franks

Abstract: These largely expository notes describe the properties of the function ${\cal R}$ which assigns a number to a $4$-tuple of distinct fixed points of an orientation preserving homeomorphism or diffeomorphism of $S^2$. These largely expository notes describe the properties of the function ${\cal R}$ which assigns a number to a $4$-tuple of distinct fixed points of an orientation preserving homeomorphism or diffeomorphism of $S^2$. △ Less

Submitted 28 December, 2014; originally announced December 2014.

Comments: 11 pages

On 1-cocycles induced by a positive definite function on a locally compact abelian group

Authors: Jordan Franks, Alain Valette

Abstract: For $\varphi$ a normalized positive definite function on a locally compact abelian group $G$, we consider on the one hand the unitary representation $π_\varphi$ associated to $\varphi$ by the GNS construction, on the other hand the probability measure $μ_\varphi$ on the Pontryagin dual $\hat{G}$ provided by Bochner's theorem. We give necessary and sufficient conditions for the vanishing of 1-cohom… ▽ More For $\varphi$ a normalized positive definite function on a locally compact abelian group $G$, we consider on the one hand the unitary representation $π_\varphi$ associated to $\varphi$ by the GNS construction, on the other hand the probability measure $μ_\varphi$ on the Pontryagin dual $\hat{G}$ provided by Bochner's theorem. We give necessary and sufficient conditions for the vanishing of 1-cohomology $H^1(G,π_\varphi)$ and reduced 1-cohomology $\bar{H}^1(G,π_\varphi)$. For example, $\bar{H}^1(G,π_\varphi)=0$ if and only if either $Hom(G,\mathbb{C})=0$ or $μ_\varphi(1_G)=0$, where $1_G$ is the trivial character of $G$. △ Less

Submitted 18 March, 2013; originally announced March 2013.

Comments: 7 pages

Some virtually abelian subgroups of the group of analytic symplectic diffeomorphisms of $S^2$

Authors: John Franks, Michael Handel

Abstract: We show that if $M$ is a compact oriented surface of genus 0 and $G$ is a subgroup of $\Symp^ω_μ(M)$ which has an infinite normal solvable subgroup, then $G$ is virtually abelian. In particular the centralizer of an infinite order $f \in \Symp^ω_μ(M)$ is virtually abelian. Another immediate corollary is that if $G$ is a solvable subgroup of $\Symp^ω_μ(M)$ then $G$ is virtually abelian. We also pro… ▽ More We show that if $M$ is a compact oriented surface of genus 0 and $G$ is a subgroup of $\Symp^ω_μ(M)$ which has an infinite normal solvable subgroup, then $G$ is virtually abelian. In particular the centralizer of an infinite order $f \in \Symp^ω_μ(M)$ is virtually abelian. Another immediate corollary is that if $G$ is a solvable subgroup of $\Symp^ω_μ(M)$ then $G$ is virtually abelian. We also prove a special case of the Tits Alternative for subgroups of $\Symp^ω_μ(S^2).$ △ Less

Submitted 8 September, 2013; v1 submitted 17 April, 2012; originally announced April 2012.

Comments: Corrected typos, references. Added new proposition (5.14) in this version

MSC Class: 37C85; 37J10

Triviality of some representations of $MCG(S_g)$ in $GL(n,C), Diff(S^2)$ and $Homeo(T^2)$

Authors: John Franks, Michael Handel

Abstract: We show the triviality of representations of the mapping class group of a genus $g$ surface in $GL(n,C), Diff(S^2)$ and $Homeo(T^2)$ when appropriate restrictions on the genus $g$ and the size of $n$ hold. For example, if $S_g$ is a surface of finite type and $φ: MCG(S_g) \to GL(n,C)$ is a homomorphism, then $φ$ is trivial provided the genus $g \ge 3$ and $n < 2g$. We also show that if $S_g$ is a… ▽ More We show the triviality of representations of the mapping class group of a genus $g$ surface in $GL(n,C), Diff(S^2)$ and $Homeo(T^2)$ when appropriate restrictions on the genus $g$ and the size of $n$ hold. For example, if $S_g$ is a surface of finite type and $φ: MCG(S_g) \to GL(n,C)$ is a homomorphism, then $φ$ is trivial provided the genus $g \ge 3$ and $n < 2g$. We also show that if $S_g$ is a closed surface with genus $g \ge 7$, then every homomorphism $φ: MCG(S_g) \to Diff(S^2)$ is trivial and that if $g \ge 3$, then every homomorphism $φ: MCG(S_g) \to Homeo(T^2)$ is trivial. △ Less

Submitted 20 April, 2011; v1 submitted 22 February, 2011; originally announced February 2011.

MSC Class: 57M60

Entropy zero area preserving diffeomorphisms of $S^2$

Authors: John Franks, Michael Handel

Abstract: In this paper we formulate and prove a structure theorem for area preserving diffeomorphisms of genus zero surfaces with zero entropy. As an application we relate the existence of faithful actions of a finite index subgroup of the mapping class group of a closed surface $Σ_g$ on $S^2$ by area preserving diffeomorphisms to the existence of finite index subgroups of bounded mapping class groups… ▽ More In this paper we formulate and prove a structure theorem for area preserving diffeomorphisms of genus zero surfaces with zero entropy. As an application we relate the existence of faithful actions of a finite index subgroup of the mapping class group of a closed surface $Σ_g$ on $S^2$ by area preserving diffeomorphisms to the existence of finite index subgroups of bounded mapping class groups $MCG(S, \partial S)$ with non-trivial first cohomology. △ Less

Submitted 28 February, 2012; v1 submitted 1 February, 2010; originally announced February 2010.

Comments: 88 pages, 4 figures

MSC Class: 37C05; 37E30; 37E45; 37J10; 57M60

Journal ref: Geom. Topol. 16 (2012) 2187-2284

Notes on Measure and Integration

Authors: John Franks

Abstract: This text grew out of notes I have used in teaching a one quarter course on integration at the advanced undergraduate level. My intent is to introduce the Lebesgue integral in a quick, and hopefully painless, way and then go on to investigate the standard convergence theorems and a brief introduction to the Hilbert space of $L^2$ functions on the interval. The actual construction of Lebesgue m… ▽ More This text grew out of notes I have used in teaching a one quarter course on integration at the advanced undergraduate level. My intent is to introduce the Lebesgue integral in a quick, and hopefully painless, way and then go on to investigate the standard convergence theorems and a brief introduction to the Hilbert space of $L^2$ functions on the interval. The actual construction of Lebesgue measure and proofs of its key properties are relegated to an appendix. Instead the text introduces Lebesgue measure as a generalization of the concept of length and motivates its key properties: monotonicity, countable additivity, and translation invariance. △ Less

Submitted 10 August, 2009; v1 submitted 27 February, 2008; originally announced February 2008.

Comments: This version corrects a few typos. An expanded version of this text has been published as "A (Terse) Introduction to Lebesgue Integration" as vol. 48 of the A.M.S. Student Mathematical Library

Global fixed points for centralizers and Morita's Theorem

Authors: John Franks, Michael Handel

Abstract: We prove a global fixed point theorem for the centralizer of a homeomorphism of the two dimensional disk $D$ that has attractor-repeller dynamics on the boundary with at least two attractors and two repellers. As one application, we show that there is a finite index subgroup of the centralizer of a pseudo-Anosov homeomorphism with infinitely many global fixed points. As another application we gi… ▽ More We prove a global fixed point theorem for the centralizer of a homeomorphism of the two dimensional disk $D$ that has attractor-repeller dynamics on the boundary with at least two attractors and two repellers. As one application, we show that there is a finite index subgroup of the centralizer of a pseudo-Anosov homeomorphism with infinitely many global fixed points. As another application we give an elementary proof of Morita's Theorem, that the mapping class group of a closed surface $S$ of genus $g$ does not lift to the group of diffeormorphisms of $S$ and we improve the lower bound for $g$ from 5 to 3. △ Less

Submitted 4 January, 2008; originally announced January 2008.

MSC Class: 37E30; 57M60; 37C25

Journal ref: Geom. Topol. 13 (2009) 87-98

Complete semi-conjugacies for psuedo-Anosov homeomorphisms

Authors: John Franks, Michael Handel

Abstract: Suppose $S$ is a surface of genus $\ge 2 $, $f: S \to S$ is a surface homeomorphism isotopic to a pseudo-Anosov map $α$ and suppose $\ti S$ is the universal cover of $S$ and $F$ and $A$ are lifts of $f$ and $α$ respectively. We show there is a semiconjugacy $Θ: \ti S \to \bar Ł^s \times \bar Ł^u$ from $F$ to $\bar A$, where $\bar Ł^s$ ($\bar Ł^u$) is the completion of the $R$-tree of leaves of t… ▽ More Suppose $S$ is a surface of genus $\ge 2 $, $f: S \to S$ is a surface homeomorphism isotopic to a pseudo-Anosov map $α$ and suppose $\ti S$ is the universal cover of $S$ and $F$ and $A$ are lifts of $f$ and $α$ respectively. We show there is a semiconjugacy $Θ: \ti S \to \bar Ł^s \times \bar Ł^u$ from $F$ to $\bar A$, where $\bar Ł^s$ ($\bar Ł^u$) is the completion of the $R$-tree of leaves of the stable (resp. unstable) foliation for $A$ and $\bar A$ is the map induced by $A$. We also generalize a result of Markovich and show that for any $g \in Homeo(S)$ which commutes with $f$ and has identity lift $G : \ti S \to \ti S$ and for any $(c,w)$ in the image of $Θ$ each component of $Θ^{-1}(c,w)$ is $G$-invariant. △ Less

Submitted 26 December, 2007; v1 submitted 18 December, 2007; originally announced December 2007.

MSC Class: 37E30; 37D20

Distortion in Groups of Circle and Surface Diffeomorphisms

Authors: John Franks

Abstract: In these lectures we consider how algebraic properties of discrete subgroups of Lie groups restrict the possible actions of those groups on surfaces. The results show a strong parallel between the possible actions of such a group on the circle $S^1$ and the measure preserving actions on surfaces. Our aim is the study of the (non)-existence of actions of lattices in a large class of non-compact… ▽ More In these lectures we consider how algebraic properties of discrete subgroups of Lie groups restrict the possible actions of those groups on surfaces. The results show a strong parallel between the possible actions of such a group on the circle $S^1$ and the measure preserving actions on surfaces. Our aim is the study of the (non)-existence of actions of lattices in a large class of non-compact Lie groups on surfaces. A definitive analysis of the analogous question for actions on $S^1$ was carried out by É. Ghys. Our approach is topological and insofar as possible we try to isolate properties of a group which provide the tools necessary for our analysis. The two key properties we consider are almost simplicity and the existence of a distortion element. Both will be defined and described in the lectures. Our techniques are almost all from low dimensional dynamics. But we are interested in how algebraic properties of a group -- commutativity, nilpotence, etc. affect the possible kinds of dynamics which can occur. For most of the results we will consider groups of diffeomorphisms which preserve a Borel probability measure. △ Less

Submitted 28 May, 2007; originally announced May 2007.

MSC Class: 37E30;57S30

Journal ref: Panoramas et Synth`eses, Soc. Math.de France {\bf 21} (2006) 35-52

Fixed Points of abelian actions

Authors: John Franks, Michael Handel, Kamlesh Parwani

Abstract: We prove that if $\F$ is an abelian group of $C^1$ diffeomorphisms isotopic to the identity of a closed surface $S$ of genus at least two then there is a common fixed point for all elements of $\F.$ We prove that if $\F$ is an abelian group of $C^1$ diffeomorphisms isotopic to the identity of a closed surface $S$ of genus at least two then there is a common fixed point for all elements of $\F.$ △ Less

Submitted 21 July, 2006; originally announced July 2006.

MSC Class: 37E30; 37C25

Fixed Points of abelian actions on $S^2$

Authors: John Franks, Michael Handel, Kamlesh Parwani

Abstract: We prove that if $F$ is a finitely generated abelian group of orientation preserving $C^1$ diffeomorphisms of $R^2$ which leaves invariant a compact set then there is a common fixed point for all elements of $F.$ We also show that if $F$ is any abelian subgroup of orientation preserving $C^1$ diffeomorphisms of $S^2$ then there is a common fixed point for all elements of a subgroup of $F$ with i… ▽ More We prove that if $F$ is a finitely generated abelian group of orientation preserving $C^1$ diffeomorphisms of $R^2$ which leaves invariant a compact set then there is a common fixed point for all elements of $F.$ We also show that if $F$ is any abelian subgroup of orientation preserving $C^1$ diffeomorphisms of $S^2$ then there is a common fixed point for all elements of a subgroup of $F$ with index at most two. △ Less

Submitted 29 December, 2005; v1 submitted 23 September, 2005; originally announced September 2005.

MSC Class: 37E30

Erratum to "Generalizations of the Poincaré-Birkhoff Theorem"

Authors: John Franks

Abstract: This is an erratum to an earlier paper, "Generalizations of the Poincaré-Birkhoff theorem." An error in the statement of one of the theorems is corrected. This is an erratum to an earlier paper, "Generalizations of the Poincaré-Birkhoff theorem." An error in the statement of one of the theorems is corrected. △ Less

Submitted 25 May, 2007; v1 submitted 13 October, 2004; originally announced October 2004.

Journal ref: Ann. of Math. (2) 164 (2006), no. 3, 1097--1098

Distortion Elements in Group actions on surfaces

Authors: John Franks, Michael Handel

Abstract: If $\G$ is a finitely generated group with generators $\{g_1,...,g_j\}$ then an infinite order element $f \in \G$ is a {\em distortion element} of $\G$ provided $\displaystyle{\liminf_{n \to \infty} |f^n|/n = 0,}$ where $|f^n|$ is the word length of $f^n$ in the generators. Let $S$ be a closed orientable surface and let $\Diff(S)_0$ denote the identity component of the group of $C^1$ diffeomorph… ▽ More If $\G$ is a finitely generated group with generators $\{g_1,...,g_j\}$ then an infinite order element $f \in \G$ is a {\em distortion element} of $\G$ provided $\displaystyle{\liminf_{n \to \infty} |f^n|/n = 0,}$ where $|f^n|$ is the word length of $f^n$ in the generators. Let $S$ be a closed orientable surface and let $\Diff(S)_0$ denote the identity component of the group of $C^1$ diffeomorphisms of $S$. Our main result shows that if $S$ has genus at least two and if $f$ is a distortion element in some finitely generated subgroup of $\Diff(S)_0$, then $\supp(μ) \subset \Fix(f)$ for every $f$-invariant Borel probability measure $μ$. Related results are proved for $S = T^2$ or $S^2$. For $μ$ a Borel probability measure on $S$, denote the group of $C^1$ diffeomorphisms that preserve $μ$ by $\Diff_μ(S)$. We give several applications of our main result showing that certain groups, including a large class of higher rank lattices, admit no homomorphisms to $\Diff_μ(S)$ with infinite image. △ Less

Submitted 3 March, 2005; v1 submitted 29 April, 2004; originally announced April 2004.

MSC Class: 37E30; 57S25

Periodic points of Hamiltonian surface diffeomorphisms

Authors: John Franks, Michael Handel

Abstract: The main result of this paper is that every non-trivial Hamiltonian diffeomorphism of a closed oriented surface of genus at least one has periodic points of arbitrarily high period. The same result is true for S^2 provided the diffeomorphism has at least three fixed points. In addition we show that up to isotopy relative to its fixed point set, every orientation preserving diffeomorphism F: S --… ▽ More The main result of this paper is that every non-trivial Hamiltonian diffeomorphism of a closed oriented surface of genus at least one has periodic points of arbitrarily high period. The same result is true for S^2 provided the diffeomorphism has at least three fixed points. In addition we show that up to isotopy relative to its fixed point set, every orientation preserving diffeomorphism F: S --> S of a closed orientable surface has a normal form. If the fixed point set is finite this is just the Thurston normal form. △ Less

Submitted 14 November, 2003; v1 submitted 24 March, 2003; originally announced March 2003.

Comments: Published by Geometry and Topology at http://www.maths.warwick.ac.uk/gt/GTVol7/paper20.abs.html

MSC Class: 37J10; 37E30

Journal ref: Geom. Topol. 7 (2003) 713-756

Rotation Numbers and Instability Sets

Authors: John Franks

Abstract: Translation and rotation numbers have played an interesting and important role in the qualitative description of various dynamical systems. In this exposition we are especially interested in applications which lead to proofs of periodic motions in various kinds of dynamics on the annulus. The applications include billiards and geodesic flows. Going beyond this simple qualitative invariant in t… ▽ More Translation and rotation numbers have played an interesting and important role in the qualitative description of various dynamical systems. In this exposition we are especially interested in applications which lead to proofs of periodic motions in various kinds of dynamics on the annulus. The applications include billiards and geodesic flows. Going beyond this simple qualitative invariant in the study of the dynamics of area preserving annulus maps, G.D. Birkhoff was led to the concept of ``regions of instability'' for twist maps. We discuss the closely related notion of instability sets for a generic area preserving surface diffeomorphism and develop their properties. △ Less

Submitted 24 March, 2003; originally announced March 2003.

Comments: This article is the written version of an invited address at the January, 2002 AMS meeting in San Diego, California

A Hölder continuous vector field tangent to many foliations

Authors: Christian Bonatti, John Franks

Abstract: We construct an example of a Hölder continuous vector field on the plane which is tangent to all foliations in a continuous family of pairwise distinct $C^1$ foliations. Given any $1 \le r <\infty,$ the construction can be done in such a way that each leaf of each foliation is the graph of a $C^r$ function from $\R$ to $\R.$ We also show the existence of a continuous vector field $X$ on $\R^2$ a… ▽ More We construct an example of a Hölder continuous vector field on the plane which is tangent to all foliations in a continuous family of pairwise distinct $C^1$ foliations. Given any $1 \le r <\infty,$ the construction can be done in such a way that each leaf of each foliation is the graph of a $C^r$ function from $\R$ to $\R.$ We also show the existence of a continuous vector field $X$ on $\R^2$ and two foliations $\cal{F}$ and $\cal{G}$ on $\R^2$ each tangent to $X$ with a dense subset $\cal E$ of $\R^2$ such that at every point $x\in \cal E$ the leaves $F_x$ and $G_x$ of the foliation $\cal{F}$ and $\cal{G}$ through $x$ are topologically transverse. △ Less

Submitted 24 March, 2003; originally announced March 2003.

Area preserving group actions on surfaces

Authors: John Franks, Michael Handel

Abstract: Suppose G is an almost simple group containing a subgroup isomorphic to the three-dimensional integer Heisenberg group. For example any finite index subgroup of SL(3,Z) is such a group. The main result of this paper is that every action of G on a closed oriented surface by area preserving diffeomorphisms factors through a finite group. Suppose G is an almost simple group containing a subgroup isomorphic to the three-dimensional integer Heisenberg group. For example any finite index subgroup of SL(3,Z) is such a group. The main result of this paper is that every action of G on a closed oriented surface by area preserving diffeomorphisms factors through a finite group. △ Less

Submitted 14 November, 2003; v1 submitted 15 March, 2002; originally announced March 2002.

Comments: Published by Geometry and Topology at http://www.maths.warwick.ac.uk/gt/GTVol7/paper21.abs.html

MSC Class: 57S25; 37E30

Journal ref: Geom. Topol. 7 (2003) 757-771

Groups of diffeomorphisms of one-manifolds, III: Nilpotent subgroups

Authors: Benson Farb, John Franks

Abstract: Plante-Thurston proved that every nilpotent subgroup of $\Diff^2(S^1)$ is abelian. One of our main results is a sharp converse: $\Diff^1(S^1)$ contains every finitely-generated, torsion-free nilpotent group. Plante-Thurston proved that every nilpotent subgroup of $\Diff^2(S^1)$ is abelian. One of our main results is a sharp converse: $\Diff^1(S^1)$ contains every finitely-generated, torsion-free nilpotent group. △ Less

Submitted 31 May, 2018; v1 submitted 13 August, 2001; originally announced August 2001.

Comments: This is a revision of a paper published in 2003. It corrects a minor mistake ( "2n-2j+2" changed to "4n-2j+2") in the proof of the Theorem 1.1. See Remark 2.1 in the current version

Journal ref: Ergodic Theory Dynamical Systems 23 (2003), no. 5, 1467--1484

Groups of homeomorphisms of one-manifolds, I: actions of nonlinear groups

Authors: Benson Farb, John Franks

Abstract: This self-contained paper is part of a series \cite{FF2,FF3} on actions by diffeomorphisms of infinite groups on compact manifolds. The two main results presented here are: 1) Any homomorphism of (almost any) mapping class group or automorphism group of a free group into $\Diff_+^r(S^1), r\geq 2$ is trivial. For r=0 Nielsen showed that in many cases nontrivial (even faithful) representations e… ▽ More This self-contained paper is part of a series \cite{FF2,FF3} on actions by diffeomorphisms of infinite groups on compact manifolds. The two main results presented here are: 1) Any homomorphism of (almost any) mapping class group or automorphism group of a free group into $\Diff_+^r(S^1), r\geq 2$ is trivial. For r=0 Nielsen showed that in many cases nontrivial (even faithful) representations exist. Somewhat weaker results are proven for finite index subgroups. 2) We construct a finitely-presented group of real-analytic diffeomorphisms of $\R$ which is not residually finite. △ Less

Submitted 11 July, 2001; v1 submitted 11 July, 2001; originally announced July 2001.

Comments: 28 pages

Groups of homeomorphisms of one-manifolds, III: Nilpotent subgroups

Authors: Benson Farb, John Franks

Abstract: This paper has been withdrawn by the authors; it will be incorporated into part I of the series (in preparation). This paper has been withdrawn by the authors; it will be incorporated into part I of the series (in preparation). △ Less

Submitted 19 February, 2001; v1 submitted 3 February, 2001; originally announced February 2001.

Group actions on one-manifolds, II: Extensions of Hölder's Theorem

Authors: Benson Farb, John Franks

Abstract: We study groups of homeomorphisms of R, each of whose elements have at most one fixed point. In particular we prove that any such group of C^2 diffeomorphisms is topologically conjugate to an affine group. We study groups of homeomorphisms of R, each of whose elements have at most one fixed point. In particular we prove that any such group of C^2 diffeomorphisms is topologically conjugate to an affine group. △ Less

Submitted 3 May, 2001; v1 submitted 3 February, 2001; originally announced February 2001.

Comments: references added, new theorem added :topological characterization of affine groups on R

Shift Equivalence and the Conley index

Authors: John Franks, David Richeson

Abstract: In this paper we introduce filtration pairs for isolated invariant sets of continuous maps. We prove the existence of filtration pairs and show that, up to shift equivalence, the induced map on the corresponding pointed space is an invariant of the isolated invariant set. Moreover, the maps defining the shift equivalence can be chosen canonically. Lastly, we define partially ordered Morse decomp… ▽ More In this paper we introduce filtration pairs for isolated invariant sets of continuous maps. We prove the existence of filtration pairs and show that, up to shift equivalence, the induced map on the corresponding pointed space is an invariant of the isolated invariant set. Moreover, the maps defining the shift equivalence can be chosen canonically. Lastly, we define partially ordered Morse decompositions and prove the existence of Morse set filtrations for such decompositions. △ Less

Submitted 29 October, 1999; originally announced October 1999.

Comments: 23 pages

MSC Class: 58F12; 58F20

Regions of instability for non-twist maps

Authors: John Franks, Patrice Le Calvez

Abstract: In this paper we consider an analog of the regions of instability for twist maps in the context of area preserving diffeomorphisms which are not twist maps. Several properties analogous to those of classical regions of instability are proved. In this paper we consider an analog of the regions of instability for twist maps in the context of area preserving diffeomorphisms which are not twist maps. Several properties analogous to those of classical regions of instability are proved. △ Less

Submitted 30 May, 2000; v1 submitted 27 October, 1999; originally announced October 1999.

Comments: 30 pages

MSC Class: 58F05; 34C35

The rotation set and periodic points for torus homeomorphisms

Authors: John Franks

Abstract: We consider the rotation set $ρ(F)$ for a lift $F$ of an area preserving homeomorphism $f: \t^2\to \t^2$, which is homotopic to the identity. The relationship between this set and the existence of periodic points for $f$ is least well understood in the case when this set is a line segment. We show that in this case if a vector $v$ lies in $ρ(F)$ and has both co-ordinates rational, then there is… ▽ More We consider the rotation set $ρ(F)$ for a lift $F$ of an area preserving homeomorphism $f: \t^2\to \t^2$, which is homotopic to the identity. The relationship between this set and the existence of periodic points for $f$ is least well understood in the case when this set is a line segment. We show that in this case if a vector $v$ lies in $ρ(F)$ and has both co-ordinates rational, then there is a periodic point $x\in \t^2$ with the property that $$\frac{F^q(x_0)-x_0}q = v$$ where $x_0\in \re^2$ is any lift of $x$ and $q$ is the least period of $x$. △ Less

Submitted 6 May, 1996; originally announced May 1996.

Report number: Dynamical Systems 5/7/96