A new method improves maximum inner product search by locally decomposing residual vectors.
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Trend · papers per month
Study quantization schemes on Kähler manifolds linking star products and BV quantizations.
Classifies star products on Lie algebroid duals and extends to projectable quantizations.
A new method for robust product Markovian quantization overcomes numerical instabilities.
New vector quantization method reduces relevance of parallel components in database points.
For arbitrary compact quantizable Kaehler manifolds it is shown how a natural formal deformation quantization (star product) can be obtained via Berezin-Toeplitz operators. Results on their semi-classical behaviour (their asymptotic expansion) due to Bordemann, Meinrenken and Schlichenmaier are used in an essential man…
We propose a quantization based approach for fast approximate Maximum Inner Product Search (MIPS). Each database vector is quantized in multiple subspaces via a set of codebooks, learned directly by minimizing the inner product quantization error. Then, the inner product of a query to a database vector is approximated …
I have chosen, in this presentation of Deformation Quantization, to focus on 3 points: the uniqueness --up to equivalence-- of a universal star product (universal in the sense of Kontsevich) on the dual of a Lie algebra, the cohomology classes introduced by Deligne for equivalence classes of differential star products …
Lectures on symplectic and Poisson geometry, quantization, and quantum field theory.
Paper discusses star products and Kähler metrics, linking deformation quantization and constant curvature metrics.
Study provides explicit formula for complex 2D Kähler manifold quantization.
Formally equates two quantization methods and constructs non-commutative algebras.
The paper constructs a star product on a symplectically reduced phase space for a lattice gauge model.
Vectors of data are at the heart of machine learning and data mining. Recently, vector quantization methods have shown great promise in reducing both the time and space costs of operating on vectors. We introduce a vector quantization algorithm that can compress vectors over 12x faster than existing techniques while al…
Improved Heston model produces steeper smile for short maturities.
We study the Berezin-Toeplitz quantization using as quantum space the space of eigenstates of the renormalized Bochner Laplacian corresponding to eigenvalues localized near the origin on a symplectic manifold. We show that this quantization has the correct semiclassical behavior and construct the corresponding star-pro…
We formulate a quantization commutes with reduction principle in the setting where the Lie group , the symplectic manifold it acts on, and the orbit space of the action may all be noncompact. It is assumed that the action is proper, and the zero set of a deformation vector field, associated to the momentum map and a…
We propose an explicit construction of the deformation quantization of the general second-class constrained system, which is covariant with respect to local coordinates on the phase space. The approach is based on constructing the effective first-class constraint (gauge) system equivalent to the original second-class o…
Quantization on even-dimensional compact manifolds using cell decomposition.
Paper proposes a method to speed up DNNs by quantizing Winograd/Toom-Cook convolutions.
Develops a new method for quantizing rough volatility for volatility derivatives pricing.
The paper constructs quantizations for symplectic manifolds with specific Laplacian properties.
The main goal of this paper is to compute the characteristic class of the Alekseev-Lachowska *-product on coadjoint orbits. We deduce an analogue of the Weyl dimension formula in the context of deformation quantization.
It is known that holomorphic Poisson structures are closely related to theories of generalized Kähler geometry and bi-Hermitian structures. In this article, we introduce quantization of holomorphic Poisson structures which are closely related to generalized Kähler structures /bi-Hermitian structures. By resulting nonco…
Quantizes symplectic manifolds with toric singularities using Toeplitz operators.
Noncommutative geometry connects higher order connections to quantization.
Kasparov defined a distinguished K-homology fundamental class, so called the Dirac element. We prove a localization formula for the Dirac element in K-homology of crossed product of C^{*}-algebras. Then we define the quantization of Hamiltonian G-spaces as a push-forward of the Dirac element. With this, we develop a K-…
This paper develops quantization algorithms for random Fourier features, simplifying the process and improving performance.
This work reduces model size by 86.11% for recommender systems using 4-bit quantization.
This work is a contribution to the area of Strict Quantization (in the sense of Rieffel) in the presence of curvature and non-Abelian group actions. More precisely, we use geometry to obtain explicit oscillatory integral formulae for strongly invariant strict deformation quantizations of a class of solvable symplectic …
Researchers create a star product on a Grassmannian with separation of variables.
The article defines and compares two types of quantizations on compact manifolds.
Paper proposes methods for pricing FX-linked Bermudan options using quantization.
RATQ is a new quantizer for optimizing noisy gradients in machine learning.
Paper quantizes heavy-tailed data for near optimal estimation rates.
A symplectic fibration is a fibre bundle in the symplectic category. We find the relation between deformation quantization of the base and the fibre, and the total space. We use the weak coupling form of Guillemin, Lerman, Sternberg and find the characteristic class of deformation of symplectic fibration. We also prove…
Kontsevich's formula for a deformation quantization of Poisson structures involves a Feynman series of graphs, with the weights given by some complicated integrals (using certain pullbacks of the standard angle form on a circe). We explain the geometric meaning of this series as degrees of maps of some grand configurat…
A new method defends neural networks from adversarial attacks.
For a real symmetric domain , with complexification , we introduce the concept of "star-restriction" (a real analogue of the "star-products" for quantization of Kähler manifolds) and give a geometric construction of the -invariant differential ope…
We investigate (pseudo)differential forms in the framework of supergeometry. Definitions, basic properties and Cartan calculus (DeRham differential, Lie derivative, inner product, Hodge operator) are presented; the symplectic supermechanics (even and odd) is formulated; and the question of quantization is discussed. In…
Verma Howe duality connects tensor products of Verma modules to LKB representations.
Paper proves existence of a universal codebook for low-precision quantization.
In this note, we will show one example of hamiltonian Lie algebra action which has no invariant star product.
DPQ offers significant compression at minimal cost for embedding layers.
It is of fundamental importance to find algorithms obtaining optimal performance for learning of statistical models in distributed and communication limited systems. Aiming at characterizing the optimal strategies, we consider learning of Gaussian Processes (GPs) in distributed systems as a pivotal example. We first ad…
DBQ quantizes lightweight networks efficiently for resource-constrained devices.
We show that the Hochschild cohomology of the algebra obtained by formal deformation quantization on a symplectic manifold is isomorphic to the formal series with coefficients in the de Rham cohomology of the manifold. The cohomology class obtained by differentiating the star-product with respect to the deformation par…
We consider formal deformations of the Poisson algebra of functions (with singularities) on which are Laurent polynomials of fibers. Tn the case: (), there exists a non-trivial -product on this algebra non-equivalent to the standard Moyal product.