Results 11 to 20 of about 412,590 (166)
Real block-circulant matrices and DCT-DST algorithm for transformer neural network
Frontiers in Applied Mathematics and Statistics, 2023 In the encoding and decoding process of transformer neural networks, a weight matrix-vector multiplication occurs in each multihead attention and feed forward sublayer.
Euis Asriani +4 more
doaj +1 more source
USE OF THE TREND-FACTOR MODEL TO IMPROVE THE ACCURACY FORECASTS
Статистика и экономика, 2016 In this paper we propose a method toimprove the accuracy of the trend-factormodel on the assumption that the increasein endogenous variable-screens dependnot only on time but also deviations fromtheir trend of exogenous variables, provedby the ...
Irina V. Orlova, Viktor B. Turundaevsky
doaj +1 more source
Matrix compression along isogenic blocks
Acta Scientiarum Mathematicarum, 2022 AbstractA matrix-compression algorithm is derived from a novel isogenic block decomposition for square matrices. The resulting compression and inflation operations possess strong functorial and spectral-permanence properties. The basic observation that Hadamard entrywise functional calculus preserves isogenic blocks has already proved to be of ...
Belton, Alexander +3 more
openaire +5 more sources
A note on the formulas for the Drazin inverse of the sum of two matrices
Open Mathematics, 2019 In this paper we derive the formula of (P + Q)D under the conditions Q(P + Q)P(P + Q) = 0, P(P + Q)P(P + Q) = 0 and QPQ2 = 0. Then, a corollary is given which satisfies the conditions (P + Q)P(P + Q) = 0 and QPQ2 = 0. Meanwhile, we show that the additive
Liu Xin, Yang Xiaoying, Wang Yaqiang
doaj +1 more source
Deep Subspace Clustering with Block Diagonal Constraint
Applied Sciences, 2020 The deep subspace clustering method, which adopts deep neural networks to learn a representation matrix for subspace clustering, has shown good performance.
Jing Liu, Yanfeng Sun, Yongli Hu
doaj +1 more source
MATRIX SUBADDITIVITY INEQUALITIES AND BLOCK-MATRICES [PDF]
International Journal of Mathematics, 2009 We give a number of subadditivity results and conjectures for symmetric norms, matrices and block-matrices. Let A, B, Z be matrices of same size and suppose that A, B are normal and Z is expansive, i.e. Z*Z ≥ I. We conjecture that [Formula: see text] for all non-negative concave function f on [0,∞) and all symmetric norms ‖ · ‖ (in particular for all ...
openaire +3 more sources
Practical method to solve large least squares problems using Cholesky decomposition
Geodesy and Cartography, 2015 In Geomatics, the method of least squares is commonly used to solve the systems of observation equations for a given number of unknowns. This method is basically implemented in case of having number observations larger than the number of unknowns ...
Ghadi Younis
doaj +1 more source
The Limiting Spectra of Girko’s Block-Matrix [PDF]
Journal of Theoretical Probability, 2007 10 ...
openaire +2 more sources
Sensors, 2015 Aiming at addressing the problem of high computational cost of the traditional Kalman filter in SINS/GPS, a practical optimization algorithm with offline-derivation and parallel processing methods based on the numerical characteristics of the system is ...
Shaoxing Hu +3 more
doaj +1 more source
Lossy Compression using Adaptive Polynomial Image Encoding
Advances in Electrical and Computer Engineering, 2021 In this paper, an efficient lossy compression approach using adaptive-block polynomial curve-fitting encoding is proposed. The main idea of polynomial curve fitting is to reduce the number of data elements in an image block to a few coefficients.
OTHMAN, S. +3 more
doaj +1 more source