Showing 1–7 of 7 results for author: Andreev, A

Real Acceleration of Communication Process in Distributed Algorithms with Compression

Authors: Svetlana Tkachenko, Artem Andreev, Aleksandr Beznosikov, Alexander Gasnikov

Abstract: Modern applied optimization problems become more and more complex every day. Due to this fact, distributed algorithms that can speed up the process of solving an optimization problem through parallelization are of great importance. The main bottleneck of distributed algorithms is communications, which can slow down the method dramatically. One way to solve this issue is to use compression of trans… ▽ More Modern applied optimization problems become more and more complex every day. Due to this fact, distributed algorithms that can speed up the process of solving an optimization problem through parallelization are of great importance. The main bottleneck of distributed algorithms is communications, which can slow down the method dramatically. One way to solve this issue is to use compression of transmitted information. In the current literature on theoretical distributed optimization, it is generally accepted that as much as we compress information, so much we reduce communication time. But in reality, the communication time depends not only on the size of the transmitted information, but also, for example, on the message startup time. In this paper, we study distributed optimization algorithms under the assumption of a more complex and closer-to-reality dependence of transmission time on compression. In particular, we describe the real speedup achieved by compression, analyze how much it makes sense to compress information, and present an adaptive way to select the power of compression depending on unknown or changing parameters of the communication process. △ Less

Submitted 10 September, 2023; originally announced September 2023.

Comments: 11 pages

Towards an architectural framework for intelligent virtual agents using probabilistic programming

Authors: Anton Andreev, Grégoire Cattan

Abstract: We present a new framework called KorraAI for conceiving and building embodied conversational agents (ECAs). Our framework models ECAs' behavior considering contextual information, for example, about environment and interaction time, and uncertain information provided by the human interaction partner. Moreover, agents built with KorraAI can show proactive behavior, as they can initiate interaction… ▽ More We present a new framework called KorraAI for conceiving and building embodied conversational agents (ECAs). Our framework models ECAs' behavior considering contextual information, for example, about environment and interaction time, and uncertain information provided by the human interaction partner. Moreover, agents built with KorraAI can show proactive behavior, as they can initiate interactions with human partners. For these purposes, KorraAI exploits probabilistic programming. Probabilistic models in KorraAI are used to model its behavior and interactions with the user. They enable adaptation to the user's preferences and a certain degree of indeterminism in the ECAs to achieve more natural behavior. Human-like internal states, such as moods, preferences, and emotions (e.g., surprise), can be modeled in KorraAI with distributions and Bayesian networks. These models can evolve over time, even without interaction with the user. ECA models are implemented as plugins and share a common interface. This enables ECA designers to focus more on the character they are modeling and less on the technical details, as well as to store and exchange ECA models. Several applications of KorraAI ECAs are possible, such as virtual sales agents, customer service agents, virtual companions, entertainers, or tutors. △ Less

Submitted 20 July, 2023; originally announced July 2023.

Abelian TQFTS and Schrödinger local systems

Authors: Aleksei Andreev, Anna Beliakova, Christian Blanchet

Abstract: We construct an action of 3-cobordisms on the finite dimensional Schrödinger representations of the Heisenberg group by Lagrangian correspondences. In addition, we review the construction of the abelian Topological Quantum Field Theory (TQFT) associated with a $q$-deformation of $U(1)$ for any root of unity $q$. We prove that for3-cobor\-disms compatible with Lagrangian correspondences, there is a… ▽ More We construct an action of 3-cobordisms on the finite dimensional Schrödinger representations of the Heisenberg group by Lagrangian correspondences. In addition, we review the construction of the abelian Topological Quantum Field Theory (TQFT) associated with a $q$-deformation of $U(1)$ for any root of unity $q$. We prove that for3-cobor\-disms compatible with Lagrangian correspondences, there is a normalization of the associated Schrödinger bimodule action that reproduces the abelian TQFT. The full abelian TQFT provides a projective representation of the mapping class group $\mathrm{Mod}(Σ)$ on the Schrödinger representation,which is linearizable at odd root of 1. Motivated by homology of surface configurations with Schrödinger representation as local coefficients, we define another projective action of $\mathrm{Mod}(Σ)$ on Schrödinger representations. We show that the latter is not linearizable by identifying the associated 2-cocycle. △ Less

Submitted 19 September, 2024; v1 submitted 19 June, 2023; originally announced June 2023.

Comments: 25 pages, 3 figures

MSC Class: 57K16; 57R56

Matrix Models, Integral Polyhedra and Toric Geometry

Authors: Aleksey Andreev

Abstract: We propose to take a look at a new approach to the study of integral polyhedra. The main idea is to give an integral representation, or matrix model representation, for the key combinatorial characteristics of integral polytopes. Based on the well-known geometric interpretations of matrix model digram techniques, we construct a new model that enumerates triangulations, subdivisions, and numbers of… ▽ More We propose to take a look at a new approach to the study of integral polyhedra. The main idea is to give an integral representation, or matrix model representation, for the key combinatorial characteristics of integral polytopes. Based on the well-known geometric interpretations of matrix model digram techniques, we construct a new model that enumerates triangulations, subdivisions, and numbers of integral points of integral polygons. This approach allows us to look at their combinatorics from a new perspective, motivated by knowledge about matrix models and their integrability. We show how analogs of Virasoro constraints appear in the resulting model. Moreover, we make a natural generalization of this matrix model to the case of polytopes of an arbitrary dimension, considering already a tensor model. We also obtain an analogue of Virasoro constraints for it and discuss their role in the solvability of these models. The deep connection between the geometry of convex polyhedra and toric geometry is the main reference point in the construction of these models. We present considerations on specific ways of applying this approach to the description of Batyrev's mirror pairs. All this allows us to formulate many interesting directions in the study of the connection between matrix/tensor models and the geometry of toric varieties. △ Less

Submitted 18 October, 2022; v1 submitted 12 October, 2022; originally announced October 2022.

Comments: 19 pages; version 2: some references added

Report number: MIPT/TH-20/22; ITEP/TH-23/22; IITP/TH-22/22

Combinatorics of KP hierarchy structural constants

Authors: A. Andreev, A. Popolitov, A. Sleptsov, A. Zhabin

Abstract: Following Natanzon-Zabrodin, we explore the Kadomtsev-Petviashvili hierarchy as an infinite system of mutually consistent relations on the second derivatives of the free energy with some universal coefficients. From this point of view, various combinatorial properties of these coefficients naturally highlight certain non-trivial properties of the KP hierarchy. Furthermore, this approach allows us… ▽ More Following Natanzon-Zabrodin, we explore the Kadomtsev-Petviashvili hierarchy as an infinite system of mutually consistent relations on the second derivatives of the free energy with some universal coefficients. From this point of view, various combinatorial properties of these coefficients naturally highlight certain non-trivial properties of the KP hierarchy. Furthermore, this approach allows us to suggest several interesting directions of the KP deformation via a deformation of these coefficients. We also construct an eigenvalue matrix model, whose correlators fully describe the universal KP coefficients, which allows us to further study their properties and generalizations. △ Less

Submitted 25 January, 2021; originally announced January 2021.

Comments: 17 pages, no figures

Report number: MIPT/TH-19/20; ITEP/TH-34/20; IITP/TH-21/20

Journal ref: Eur. Phys. J. C 81, 1136 (2021)

Genus expansion of matrix models and $\hbar$ expansion of KP hierarchy

Authors: A. Andreev, A. Popolitov, A. Sleptsov, A. Zhabin

Abstract: We study $\hbar$ expansion of the KP hierarchy following Takasaki-Takebe arXiv:hep-th/9405096 considering several examples of matrix model $τ$-functions with natural genus expansion. Among the examples there are solutions of KP equations of special interest, such as generating function for simple Hurwitz numbers, Hermitian matrix model, Kontsevich model and Brezin-Gross-Witten model. We show that… ▽ More We study $\hbar$ expansion of the KP hierarchy following Takasaki-Takebe arXiv:hep-th/9405096 considering several examples of matrix model $τ$-functions with natural genus expansion. Among the examples there are solutions of KP equations of special interest, such as generating function for simple Hurwitz numbers, Hermitian matrix model, Kontsevich model and Brezin-Gross-Witten model. We show that all these models with parameter $\hbar$ are $τ$-functions of the $\hbar$-KP hierarchy and the expansion in $\hbar$ for the $\hbar$-KP coincides with the genus expansion for these models. Furthermore, we show a connection of recent papers considering the $\hbar$-formulation of the KP hierarchy arXiv:1509.04472, arXiv:1512.07172 with original Takasaki-Takebe approach. We find that in this approach the recovery of enumerative geometric meaning of $τ$-functions is straightforward and algorithmic. △ Less

Submitted 7 September, 2020; v1 submitted 14 August, 2020; originally announced August 2020.

Comments: 26 pages, no figures

Report number: ITEP/TH-15/20; IITP/TH-11/20; MIPT/TH-10/20

Journal ref: J. High Energ. Phys. 2020, 38 (2020)

A New Generation of Brain-Computer Interface Based on Riemannian Geometry

Authors: Marco Congedo, Alexandre Barachant, Anton Andreev

Abstract: Based on the cumulated experience over the past 25 years in the field of Brain-Computer Interface (BCI) we can now envision a new generation of BCI. Such BCIs will not require training; instead they will be smartly initialized using remote massive databases and will adapt to the user fast and effectively in the first minute of use. They will be reliable, robust and will maintain good performances… ▽ More Based on the cumulated experience over the past 25 years in the field of Brain-Computer Interface (BCI) we can now envision a new generation of BCI. Such BCIs will not require training; instead they will be smartly initialized using remote massive databases and will adapt to the user fast and effectively in the first minute of use. They will be reliable, robust and will maintain good performances within and across sessions. A general classification framework based on recent advances in Riemannian geometry and possessing these characteristics is presented. It applies equally well to BCI based on event-related potentials (ERP), sensorimotor (mu) rhythms and steady-state evoked potential (SSEP). The framework is very simple, both algorithmically and computationally. Due to its simplicity, its ability to learn rapidly (with little training data) and its good across-subject and across-session generalization, this strategy a very good candidate for building a new generation of BCIs, thus we hereby propose it as a benchmark method for the field. △ Less

Submitted 30 October, 2013; originally announced October 2013.

Comments: 33 pages, 9 Figures, 17 equations/algorithms