-
Proceedings of the 13th International Workshop on Non-Classical Models of Automata and Applications
Authors:
Benedek Nagy,
Rudolf Freund
Abstract:
The Thirteenth International Workshop on Non-Classical Models of Automata and Applications (NCMA 2023) was held in Famagusta, North Cyprus, on September 18 and 19, 2023, organized by the Eastern Mediterranean University. The NCMA workshop series was established in 2009 as an annual event for researchers working on non-classical and classical models of automata, grammars or related devices. Such mo…
▽ More
The Thirteenth International Workshop on Non-Classical Models of Automata and Applications (NCMA 2023) was held in Famagusta, North Cyprus, on September 18 and 19, 2023, organized by the Eastern Mediterranean University. The NCMA workshop series was established in 2009 as an annual event for researchers working on non-classical and classical models of automata, grammars or related devices. Such models are investigated both as theoretical models and as formal models for applications from various points of view.
△ Less
Submitted 13 September, 2023;
originally announced September 2023.
-
A P Systems Variant for Reasoning about Sequential Controllability of Boolean Networks
Authors:
Artiom Alhazov,
Vincent Ferrari-Dominguez,
Rudolf Freund,
Nicolas Glade,
Sergiu Ivanov
Abstract:
A Boolean network is a discrete dynamical system operating on vectors of Boolean variables. The action of a Boolean network can be conveniently expressed as a system of Boolean update functions, computing the new values for each component of the Boolean vector as a function of the other components. Boolean networks are widely used in modelling biological systems that can be seen as consisting of e…
▽ More
A Boolean network is a discrete dynamical system operating on vectors of Boolean variables. The action of a Boolean network can be conveniently expressed as a system of Boolean update functions, computing the new values for each component of the Boolean vector as a function of the other components. Boolean networks are widely used in modelling biological systems that can be seen as consisting of entities which can be activated or deactivated, expressed or inhibited, on or off. P systems on the other hand are classically introduced as a model of hierarchical multiset rewriting. However, over the years the community has proposed a wide range of P system variants including diverse ingredients suited for various needs. In this work, we propose a new variant -- Boolean P systems -- specifically designed for reasoning about sequential controllability of Boolean networks, and use it to first establish a crisp formalization of the problem, and then to prove that the problem of sequential controllability is PSPACE-complete. We further claim that Boolean P systems are a demonstration of how P systems can be used to construct ad hoc formalisms, custom-tailored for reasoning about specific problems, and providing new advantageous points of view.
△ Less
Submitted 28 February, 2023;
originally announced March 2023.
-
P versus B: P Systems as a Formal Framework for Controllability of Boolean Networks
Authors:
Artiom Alhazov,
Rudolf Freund,
Sergiu Ivanov
Abstract:
Membrane computing and P systems are a paradigm of massively parallel natural computing introduced by Gheorghe Păun in 1999, inspired by the structure of the living cell and by its biochemical reactions. In spite of this explicit biological motivation, P systems have not been extensively used in modelling real-world systems. To confirm this intuition, we establish a state of the art investigatio…
▽ More
Membrane computing and P systems are a paradigm of massively parallel natural computing introduced by Gheorghe Păun in 1999, inspired by the structure of the living cell and by its biochemical reactions. In spite of this explicit biological motivation, P systems have not been extensively used in modelling real-world systems. To confirm this intuition, we establish a state of the art investigation comparing the use of P systems to that of Boolean networks in this line of research. We then propose to use P systems as a tool for setting up formal frameworks to reason about other formalisms, and we introduce Boolean P systems, specifically tailored for capturing sequential controllability of Boolean networks. We show how to tackle some technical challenges and prove that sequential controllability properly embeds in the framework of Boolean P systems.
△ Less
Submitted 31 August, 2022;
originally announced August 2022.
-
Using Taylor-Approximated Gradients to Improve the Frank-Wolfe Method for Empirical Risk Minimization
Authors:
Zikai Xiong,
Robert M. Freund
Abstract:
The Frank-Wolfe method has become increasingly useful in statistical and machine learning applications, due to the structure-inducing properties of the iterates, and especially in settings where linear minimization over the feasible set is more computationally efficient than projection. In the setting of Empirical Risk Minimization -- one of the fundamental optimization problems in statistical and…
▽ More
The Frank-Wolfe method has become increasingly useful in statistical and machine learning applications, due to the structure-inducing properties of the iterates, and especially in settings where linear minimization over the feasible set is more computationally efficient than projection. In the setting of Empirical Risk Minimization -- one of the fundamental optimization problems in statistical and machine learning -- the computational effectiveness of Frank-Wolfe methods typically grows linearly in the number of data observations $n$. This is in stark contrast to the case for typical stochastic projection methods. In order to reduce this dependence on $n$, we look to second-order smoothness of typical smooth loss functions (least squares loss and logistic loss, for example) and we propose amending the Frank-Wolfe method with Taylor series-approximated gradients, including variants for both deterministic and stochastic settings. Compared with current state-of-the-art methods in the regime where the optimality tolerance $\varepsilon$ is sufficiently small, our methods are able to simultaneously reduce the dependence on large $n$ while obtaining optimal convergence rates of Frank-Wolfe methods, in both the convex and non-convex settings. We also propose a novel adaptive step-size approach for which we have computational guarantees. Last of all, we present computational experiments which show that our methods exhibit very significant speed-ups over existing methods on real-world datasets for both convex and non-convex binary classification problems.
△ Less
Submitted 21 November, 2023; v1 submitted 29 August, 2022;
originally announced August 2022.
-
Demonstration of latency-aware 5G network slicing on optical metro networks
Authors:
B. Shariati,
L. Velasco,
J. -J. Pedreño-Manresa,
A. Dochhan,
R. Casellas,
A. Muqaddas,
O. González de Dios,
L. Luque Canto,
B. Lent,
J. E. López de Vergara,
S. López-Buedo,
F. Moreno,
P. Pavón,
M. Ruiz,
S. K. Patri,
A. Giorgetti,
F. Cugini,
A. Sgambelluri,
R. Nejabati,
D. Simeonidou,
R. -P. Braun,
A. Autenrieth,
J. -P. Elbers,
J. K. Fischer,
R. Freund
Abstract:
The H2020 METRO-HAUL European project has architected a latency-aware, cost-effective, agile, and programmable optical metro network. This includes the design of semidisaggregated metro nodes with compute and storage capabilities, which interface effectively with both 5G access and multi-Tbit/s elastic optical networks in the core. In this paper, we report the automated deployment of 5G services,…
▽ More
The H2020 METRO-HAUL European project has architected a latency-aware, cost-effective, agile, and programmable optical metro network. This includes the design of semidisaggregated metro nodes with compute and storage capabilities, which interface effectively with both 5G access and multi-Tbit/s elastic optical networks in the core. In this paper, we report the automated deployment of 5G services, in particular, a public safety video surveillance use case employing low-latency object detection and tracking using on-camera and on-the-edge analytics. The demonstration features flexible deployment of network slice instances, implemented in terms of European Telecommunications Standards Institute (ETSI) network function virtualization network services. We summarize the key findings in a detailed analysis of end-to-end quality of service, service setup time, and soft-failure detection time. The results show that the round-trip time over an 80 km link is under 800s and the service deployment time is under 180s.
△ Less
Submitted 21 February, 2022;
originally announced February 2022.
-
A Latency-Aware Real-Time Video Surveillance Demo: Network Slicing for Improving Public Safety
Authors:
B. Shariati,
J. J. Pedreno-Manresa,
A. Dochhan,
A. S. Muqaddas,
R. Casellas,
O. González de Dios,
L. L. Canto,
B. Lent,
J. E. López de Vergara,
S. López-Buedo,
F. J. Moreno,
P. Pavón,
L. Velasco,
S. Patri,
A. Giorgetti,
F. Cugini,
A. Sgambelluri,
R. Nejabati,
D. Simeonidou,
R,
-P,
Braun,
A. Autenrieth,
J. -P. Elbers,
J. K. Fischer
, et al. (1 additional authors not shown)
Abstract:
We report the automated deployment of 5G services across a latency-aware, semidisaggregated, and virtualized metro network. We summarize the key findings in a detailed analysis of end-to-end latency, service setup time, and soft-failure detection time.
We report the automated deployment of 5G services across a latency-aware, semidisaggregated, and virtualized metro network. We summarize the key findings in a detailed analysis of end-to-end latency, service setup time, and soft-failure detection time.
△ Less
Submitted 6 July, 2021;
originally announced July 2021.
-
All-Indoor Optical Customer Premises Equipment for Fixed Wireless Access
Authors:
Dominic Schulz,
Julian Hohmann,
Peter Hellwig,
Christoph Kottke,
Ronald Freund,
Volker Jungnickel,
Ralf-Peter Braun,
Frank Geilhardt
Abstract:
We demonstrate an LED-based optical wireless link for fixed wireless access applications, at data rates of 1.5 Gbit/s over 50 m. Transmission between indoor equipment and outdoor access point is possible through metal-coated insulation glass.
We demonstrate an LED-based optical wireless link for fixed wireless access applications, at data rates of 1.5 Gbit/s over 50 m. Transmission between indoor equipment and outdoor access point is possible through metal-coated insulation glass.
△ Less
Submitted 26 May, 2021;
originally announced July 2021.
-
Stochastic Frank-Wolfe for Constrained Finite-Sum Minimization
Authors:
Geoffrey Négiar,
Gideon Dresdner,
Alicia Tsai,
Laurent El Ghaoui,
Francesco Locatello,
Robert M. Freund,
Fabian Pedregosa
Abstract:
We propose a novel Stochastic Frank-Wolfe (a.k.a. conditional gradient) algorithm for constrained smooth finite-sum minimization with a generalized linear prediction/structure. This class of problems includes empirical risk minimization with sparse, low-rank, or other structured constraints. The proposed method is simple to implement, does not require step-size tuning, and has a constant per-itera…
▽ More
We propose a novel Stochastic Frank-Wolfe (a.k.a. conditional gradient) algorithm for constrained smooth finite-sum minimization with a generalized linear prediction/structure. This class of problems includes empirical risk minimization with sparse, low-rank, or other structured constraints. The proposed method is simple to implement, does not require step-size tuning, and has a constant per-iteration cost that is independent of the dataset size. Furthermore, as a byproduct of the method we obtain a stochastic estimator of the Frank-Wolfe gap that can be used as a stop** criterion. Depending on the setting, the proposed method matches or improves on the best computational guarantees for Stochastic Frank-Wolfe algorithms. Benchmarks on several datasets highlight different regimes in which the proposed method exhibits a faster empirical convergence than related methods. Finally, we provide an implementation of all considered methods in an open-source package.
△ Less
Submitted 8 September, 2022; v1 submitted 26 February, 2020;
originally announced February 2020.
-
Condition Number Analysis of Logistic Regression, and its Implications for Standard First-Order Solution Methods
Authors:
Robert M. Freund,
Paul Grigas,
Rahul Mazumder
Abstract:
Logistic regression is one of the most popular methods in binary classification, wherein estimation of model parameters is carried out by solving the maximum likelihood (ML) optimization problem, and the ML estimator is defined to be the optimal solution of this problem. It is well known that the ML estimator exists when the data is non-separable, but fails to exist when the data is separable. Fir…
▽ More
Logistic regression is one of the most popular methods in binary classification, wherein estimation of model parameters is carried out by solving the maximum likelihood (ML) optimization problem, and the ML estimator is defined to be the optimal solution of this problem. It is well known that the ML estimator exists when the data is non-separable, but fails to exist when the data is separable. First-order methods are the algorithms of choice for solving large-scale instances of the logistic regression problem. In this paper, we introduce a pair of condition numbers that measure the degree of non-separability or separability of a given dataset in the setting of binary classification, and we study how these condition numbers relate to and inform the properties and the convergence guarantees of first-order methods. When the training data is non-separable, we show that the degree of non-separability naturally enters the analysis and informs the properties and convergence guarantees of two standard first-order methods: steepest descent (for any given norm) and stochastic gradient descent. Expanding on the work of Bach, we also show how the degree of non-separability enters into the analysis of linear convergence of steepest descent (without needing strong convexity), as well as the adaptive convergence of stochastic gradient descent. When the training data is separable, first-order methods rather curiously have good empirical success, which is not well understood in theory. In the case of separable data, we demonstrate how the degree of separability enters into the analysis of $\ell_2$ steepest descent and stochastic gradient descent for delivering approximate-maximum-margin solutions with associated computational guarantees as well. This suggests that first-order methods can lead to statistically meaningful solutions in the separable case, even though the ML solution does not exist.
△ Less
Submitted 19 October, 2018;
originally announced October 2018.
-
(Tissue) P Systems with Vesicles of Multisets
Authors:
Artiom Alhazov,
Rudolf Freund,
Sergiu Ivanov,
Sergey Verlan
Abstract:
We consider tissue P systems working on vesicles of multisets with the very simple operations of insertion, deletion, and substitution of single objects. With the whole multiset being enclosed in a vesicle, sending it to a target cell can be indicated in those simple rules working on the multiset. As derivation modes we consider the sequential mode, where exactly one rule is applied in a derivatio…
▽ More
We consider tissue P systems working on vesicles of multisets with the very simple operations of insertion, deletion, and substitution of single objects. With the whole multiset being enclosed in a vesicle, sending it to a target cell can be indicated in those simple rules working on the multiset. As derivation modes we consider the sequential mode, where exactly one rule is applied in a derivation step, and the set maximal mode, where in each derivation step a non-extendable set of rules is applied. With the set maximal mode, computational completeness can already be obtained with tissue P systems having a tree structure, whereas tissue P systems even with an arbitrary communication structure are not computationally complete when working in the sequential mode. Adding polarizations (-1, 0, 1 are sufficient) allows for obtaining computational completeness even for tissue P systems working in the sequential mode.
△ Less
Submitted 21 August, 2017;
originally announced August 2017.
-
A New Perspective on Boosting in Linear Regression via Subgradient Optimization and Relatives
Authors:
Robert M. Freund,
Paul Grigas,
Rahul Mazumder
Abstract:
In this paper we analyze boosting algorithms in linear regression from a new perspective: that of modern first-order methods in convex optimization. We show that classic boosting algorithms in linear regression, namely the incremental forward stagewise algorithm (FS$_\varepsilon$) and least squares boosting (LS-Boost($\varepsilon$)), can be viewed as subgradient descent to minimize the loss functi…
▽ More
In this paper we analyze boosting algorithms in linear regression from a new perspective: that of modern first-order methods in convex optimization. We show that classic boosting algorithms in linear regression, namely the incremental forward stagewise algorithm (FS$_\varepsilon$) and least squares boosting (LS-Boost($\varepsilon$)), can be viewed as subgradient descent to minimize the loss function defined as the maximum absolute correlation between the features and residuals. We also propose a modification of FS$_\varepsilon$ that yields an algorithm for the Lasso, and that may be easily extended to an algorithm that computes the Lasso path for different values of the regularization parameter. Furthermore, we show that these new algorithms for the Lasso may also be interpreted as the same master algorithm (subgradient descent), applied to a regularized version of the maximum absolute correlation loss function. We derive novel, comprehensive computational guarantees for several boosting algorithms in linear regression (including LS-Boost($\varepsilon$) and FS$_\varepsilon$) by using techniques of modern first-order methods in convex optimization. Our computational guarantees inform us about the statistical properties of boosting algorithms. In particular they provide, for the first time, a precise theoretical description of the amount of data-fidelity and regularization imparted by running a boosting algorithm with a prespecified learning rate for a fixed but arbitrary number of iterations, for any dataset.
△ Less
Submitted 16 May, 2015;
originally announced May 2015.
-
Cooperating Distributed Grammar Systems of Finite Index Working in Hybrid Modes
Authors:
Henning Fernau,
Rudolf Freund,
Markus Holzer
Abstract:
We study cooperating distributed grammar systems working in hybrid modes in connection with the finite index restriction in two different ways: firstly, we investigate cooperating distributed grammar systems working in hybrid modes which characterize programmed grammars with the finite index restriction; looking at the number of components of such systems, we obtain surprisingly rich lattice struc…
▽ More
We study cooperating distributed grammar systems working in hybrid modes in connection with the finite index restriction in two different ways: firstly, we investigate cooperating distributed grammar systems working in hybrid modes which characterize programmed grammars with the finite index restriction; looking at the number of components of such systems, we obtain surprisingly rich lattice structures for the inclusion relations between the corresponding language families. Secondly, we impose the finite index restriction on cooperating distributed grammar systems working in hybrid modes themselves, which leads us to new characterizations of programmed grammars of finite index.
△ Less
Submitted 21 May, 2014;
originally announced May 2014.
-
One-dimensional Array Grammars and P Systems with Array Insertion and Deletion Rules
Authors:
Rudolf Freund,
Sergiu Ivanov,
Marion Oswald,
K. G. Subramanian
Abstract:
We consider the (one-dimensional) array counterpart of contextual as well as insertion and deletion string grammars and consider the operations of array insertion and deletion in array grammars. First we show that the emptiness problem for P systems with (one-dimensional) insertion rules is undecidable. Then we show computational completeness of P systems using (one-dimensional) array insertion an…
▽ More
We consider the (one-dimensional) array counterpart of contextual as well as insertion and deletion string grammars and consider the operations of array insertion and deletion in array grammars. First we show that the emptiness problem for P systems with (one-dimensional) insertion rules is undecidable. Then we show computational completeness of P systems using (one-dimensional) array insertion and deletion rules even of norm one only. The main result of the paper exhibits computational completeness of one-dimensional array grammars using array insertion and deletion rules of norm at most two.
△ Less
Submitted 5 September, 2013;
originally announced September 2013.
-
How to Obtain Computational Completeness in P Systems with One Catalyst
Authors:
Rudolf Freund,
Gheorghe Păun
Abstract:
Whether P systems with only one catalyst can already be computationally complete, is still an open problem. Here we establish computational completeness by using specific variants of additional control mechanisms. At each step using only multiset rewriting rules from one set of a finite number of sets of multiset rewriting rules allows for obtaining computational completeness with one catalyst and…
▽ More
Whether P systems with only one catalyst can already be computationally complete, is still an open problem. Here we establish computational completeness by using specific variants of additional control mechanisms. At each step using only multiset rewriting rules from one set of a finite number of sets of multiset rewriting rules allows for obtaining computational completeness with one catalyst and only one membrane. If the targets are used for choosing the multiset of rules to be applied, for getting computational completeness with only one catalyst more than one membrane is needed. If the available sets of rules change periodically with time, computational completeness can be obtained with one catalyst in one membrane. Moreover, we also improve existing computational completeness results for P systems with mobile catalysts and for P systems with membrane creation.
△ Less
Submitted 5 September, 2013;
originally announced September 2013.
-
AdaBoost and Forward Stagewise Regression are First-Order Convex Optimization Methods
Authors:
Robert M. Freund,
Paul Grigas,
Rahul Mazumder
Abstract:
Boosting methods are highly popular and effective supervised learning methods which combine weak learners into a single accurate model with good statistical performance. In this paper, we analyze two well-known boosting methods, AdaBoost and Incremental Forward Stagewise Regression (FS$_\varepsilon$), by establishing their precise connections to the Mirror Descent algorithm, which is a first-order…
▽ More
Boosting methods are highly popular and effective supervised learning methods which combine weak learners into a single accurate model with good statistical performance. In this paper, we analyze two well-known boosting methods, AdaBoost and Incremental Forward Stagewise Regression (FS$_\varepsilon$), by establishing their precise connections to the Mirror Descent algorithm, which is a first-order method in convex optimization. As a consequence of these connections we obtain novel computational guarantees for these boosting methods. In particular, we characterize convergence bounds of AdaBoost, related to both the margin and log-exponential loss function, for any step-size sequence. Furthermore, this paper presents, for the first time, precise computational complexity results for FS$_\varepsilon$.
△ Less
Submitted 3 July, 2013;
originally announced July 2013.
-
Graph-Controlled Insertion-Deletion Systems
Authors:
Rudolf Freund,
Marian Kogler,
Yurii Rogozhin,
Sergey Verlan
Abstract:
In this article, we consider the operations of insertion and deletion working in a graph-controlled manner. We show that like in the case of context-free productions, the computational power is strictly increased when using a control graph: computational completeness can be obtained by systems with insertion or deletion rules involving at most two symbols in a contextual or in a context-free manne…
▽ More
In this article, we consider the operations of insertion and deletion working in a graph-controlled manner. We show that like in the case of context-free productions, the computational power is strictly increased when using a control graph: computational completeness can be obtained by systems with insertion or deletion rules involving at most two symbols in a contextual or in a context-free manner and with the control graph having only four nodes.
△ Less
Submitted 10 August, 2010;
originally announced August 2010.
-
Drip and Mate Operations Acting in Test Tube Systems and Tissue-like P systems
Authors:
Rudolf Freund,
Marian Kogler
Abstract:
The operations drip and mate considered in (mem)brane computing resemble the operations cut and recombination well known from DNA computing. We here consider sets of vesicles with multisets of objects on their outside membrane interacting by drip and mate in two different setups: in test tube systems, the vesicles may pass from one tube to another one provided they fulfill specific constraints;…
▽ More
The operations drip and mate considered in (mem)brane computing resemble the operations cut and recombination well known from DNA computing. We here consider sets of vesicles with multisets of objects on their outside membrane interacting by drip and mate in two different setups: in test tube systems, the vesicles may pass from one tube to another one provided they fulfill specific constraints; in tissue-like P systems, the vesicles are immediately passed to specified cells after having undergone a drip or mate operation. In both variants, computational completeness can be obtained, yet with different constraints for the drip and mate operations.
△ Less
Submitted 25 November, 2009;
originally announced November 2009.
-
On the Number of Membranes in Unary P Systems
Authors:
Rudolf Freund,
Andreas Klein,
Martin Kutrib
Abstract:
We consider P systems with a linear membrane structure working on objects over a unary alphabet using sets of rules resembling homomorphisms. Such a restricted variant of P systems allows for a unique minimal representation of the generated unary language and in that way for an effective solution of the equivalence problem. Moreover, we examine the descriptional complexity of unary P systems wit…
▽ More
We consider P systems with a linear membrane structure working on objects over a unary alphabet using sets of rules resembling homomorphisms. Such a restricted variant of P systems allows for a unique minimal representation of the generated unary language and in that way for an effective solution of the equivalence problem. Moreover, we examine the descriptional complexity of unary P systems with respect to the number of membranes.
△ Less
Submitted 29 July, 2009;
originally announced July 2009.
