检索结果-内蒙古大学图书馆

One Self-Configurable Model to Solve Many Abstract Visual Reasoning Problems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Malkiński, Mikolaj Mańdziuk, Jacek Faculty of Mathematics and Information Science Warsaw University of Technology Warsaw Poland Faculty of Computer Science AGH University of Krakow Krakow Poland

Visual Reasoning (AVR) comprises a wide selection of various problems similar to those used in human IQ tests. Recent years have brought dynamic progress in solving particular AVR tasks, however, in the contemporary literature AVR problems are largely dealt with in isolation, leading to highly specialized task-specific methods. With the aim of developing universal learning systems in the AVR domain, we propose the unified model for solving Single-Choice Abstract visual Reasoning tasks (SCAR), capable of solving various single-choice AVR tasks, without making any a priori assumptions about the task structure, in particular the number and location of panels. The proposed model relies on a novel Structure-Aware dynamic Layer (SAL), which adapts its weights to the structure of the considered AVR problem. Experiments conducted on Raven’s Progressive Matrices, Visual Analogy Problems, and Odd One Out problems show that SCAR (SAL-based models, in general) effectively solves diverse AVR tasks, and its performance is on par with the state-of-the-art task-specific baselines. What is more, SCAR demonstrates effective knowledge reuse in multi-task and transfer learning settings. To our knowledge, this work is the first successful attempt to construct a general single-choice AVR solver relying on self-configurable architecture and unified solving method. With this work we aim to stimulate and foster progress on task-independent research paths in the AVR domain, with the long-term goal of development of a general AVR solver. Copyright © 2023, The Authors. All rights reserved.

关键词： Learning systems

Improving Optimization With Gaussian Processes in the Covariance Matrix Adaptation Evolution Strategy 23

学校读者我要写书评

暂无评论

Improving Optimization With Gaussian Processes in the Covari...

23rd Conference Information Technologies - Applications and Theory, ITAT 2023

作者： Tumpach, Jiří Koza, Jan Holeňa, Martin Charles University Faculty of Mathematics and Physics Prague Czech Republic Czech Technical University Faculty of Information Technology Prague Czech Republic Czech Academy of Sciences Institute of Computer Science Prague Czech Republic

This paper explores the use of Gaussian processes (GPs) in the covariance matrix adaptation evolution strategy (CMA-ES) for black-box optimization. GPs are powerful probabilistic models that capture complex relationships, making them suitable for modeling uncertain objective functions. Integrating GPs into the CMA-ES improves exploration and adaptation in the search space, enhancing convergence speed and solution quality. The paper describes a novel implementation framework allowing to use GPs as surrogate models for the CMA-ES. That framework findings encourage further research to advance the application of GPs in black-box optimization. © 2023 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (***)

关键词： Optimization

Spin-resolved nonlocal transport in proximitized Rashba nanowires

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Szumniak, Pawel Loss, Daniel Klinovaja, Jelena AGH University of Krakow Faculty of Physics and Applied Computer Science Al. Mickiewicza 30 Kraków30-059 Poland Department of Physics University of Basel Klingelbergstrasse 82 Basel4056 Switzerland

Non-equilibrium transport in hybrid semiconductor-superconductor nanowires is crucial for many quantum phenomena such as generating entangled states via cross Andreev reflection (CAR) processes, detecting topological superconductivity, reading out Andreev spin qubits, coupling spin qubits over long distances and so on. Here, we investigate numerically transport properties of a proximitized Rashba nanowire that hosts spin-polarized low-energy quasiparticle states. We show that the spin polarization in such one-dimensional Andreev bands, extended over the entire nanowire length, can be detected in nonlocal transport measurements with tunnel-coupled side leads that are spin polarized. Remarkably, we find an exact correspondence between the sign of the nonlocal conductance and the spin density of the superconducting quasiparticles at the side lead position. We demonstrate that this feature is robust to moderate static disorder. As an example, we show that such a method can be used to detect spin inversion of the bands, accompanying the topological phase transition (TPT) for realistic system parameters. Furthermore, we show that such effects can be used to switch between CAR and elastic cotunneling (ECT) processes by tuning the strength of either the electric or the magnetic field. These findings hold significant practical implications for state-of-the-art transport experiments in such hybrid systems. Copyright © 2024, The Authors. All rights reserved.

关键词： Nanowires

Prospect study for measurement of Hbb¯ coupling at the LHC and FCC-hh

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Abdy, Zahra Najafabadi, Mojtaba Mohammadi AGH University of Science and Technology Faculty of Physics and Applied Computer Science Al. Mickiewicza 30 Krakow30-055 Poland Experimental Physics Department CERN Geneva 231211 Switzerland P.O. Box 19395-5531 Tehran Iran

This paper employs the H + b + jets signature in proton-proton collisions to explore the structure of the Hbb¯ *** focus of the analysis lies in the decay of the Higgs boson into a photon pair, taking into account both reducible and irreducible backgrounds and a realistic simulation of the detector effects. To enhance the discrimination between signal and background, a multivariate analysis is employed to analyse the kinematic variables and optimise the signal-to-background ratio. The results indicate that the H + b+ jets process can significantly contribute to the precise measurement of CP-even and CP-odd couplings between the bottom quark and the Higgs boson at the LHC and FCC-hh. Finally, a novel asymmetry is introduced for the purpose of probing CP violation within the Hbb¯ coupling, formulated exclusively based on lab-frame momenta. © 2024, CC BY.

关键词： Couplings

Upper limit to quantum interaction strength between free electrons and electromagnetic single modes

学校读者我要写书评

暂无评论

Upper limit to quantum interaction strength between free ele...

CLEO: Fundamental science, CLEO:FS 2024 - Part of Conference on Lasers and Electro-Optics, CLEO 2024

作者： Xie, Zetao Chen, Zeling Li, Hao Yan, Qinghui Chen, Hongsheng Lin, Xiao Kaminer, Ido Miller, Owen D. Yang, Yi Department of Physics HK Institute of Quantum Science and Technology The University of Hong Kong Pokfulam Hong Kong Department of Applied Physics Energy Sciences Institute Yale University New HavenCT06520 United States Interdisciplinary Center for Quantum Information State Key Laboratory of Modern Optical Instrumentation ZJU-Hangzhou Global Science and Technology Innovation Center College of Information Science and Electronic Engineering Zhejiang University Hangzhou310027 China Solid State Institute Faculty of Electrical and Computer Engineering Technion - Israel Institute of Technology Haifa32000 Israel

ISBN: (纸本)9781957171395

Free-electron quantum optics facilitates electron-photon entanglement for quantum information processing. The challenge is finding specialized photonic-structures optimizing electron-photon interactions. We present a first-principle upper limit on the interaction strength, guiding the design of future structures. © Optica Publishing Group 2024, © 2024 The Author(s)

关键词： Quantum optics

A Proposed Fully Transparent, Flexible, and Compact Modeled Low-Voltage TFT for Implementation of Full Adder and Subtractor

学校读者我要写书评

暂无评论

IEEE Journal on Flexible Electronics

IEEE Journal on Flexible Electronics 2024年第11期3卷 477-483页

作者： Singh Mehrolia, Mukuljeet Verma, Ankit Kumar Singh, Abhishek Chourasia, Nitesh K. Pandey, Amritanshu Rajiv Gandhi Institute of Petroleum Technology Department of Electrical and Electronics Engineering Jais Uttar Pradesh Amethi229304 India SRM Institute of Science and Technology Delhi-NCR Campus Faculty of Engineering and Technology Department of Electronics and Communication Engineering Modinagar Ghaziabad201204 India Shree Guru Gobind Singh Tricentenary University Faculty of Applied and Basic Sciences Department of Physics Haryana Gurugram122505 India Department of Electronics Engineering Varanasi221005 India

In this article, a low-voltage, fully transparent, flexible thin-film transistor (TFT) is simulated using the Silvaco-Atlas tool, in which amorphous indium-gallium–zinc oxide (a-IGZO) and HfO2 are used as the active layers, gate dielectrics of the TFT, and indium-tin oxide (ITO) serves as the electrodes for source, drain, and gate contacts. In addition, this TFT is compactly modeled and used in the implementation of full adder and full subtractor circuits through the Silvaco-Techmodeler and Silvaco-Gateway tools. The technology computer-aided design (TCAD) simulated device has an operating voltage of 2 V and good performance parameters, such as a low threshold voltage of 0.104 V, high mobility (~8.9 cm2/V-sec), high ION/ IOFF~105} , and a low subthreshold slope (SS) of 65 mV/decade. The Silvaco-Gateway tool is used for the execution of full adder and subtractor circuits. For all eight inputs (from 000 to 111), it gives quite acceptable transient characteristics in analyzing corresponding outputs as sum, carry, difference, and borrow. Silvaco-Techmodeler tools help in the compact modeling of simulated devices and generate a high level of accuracy of ~100% with very minimal error between simulated and modeled data (0.41% and 0.94%, respectively). This simulated, compactly modeled TFT would be used in the near future for analyzing complex analog and digital circuits. © 2024 IEEE.

关键词： Thin film transistors

Fine-tuning of mechanical properties in a Zn-Ag-Mg alloy via cold plastic deformation process and post-deformation annealing

学校读者我要写书评

暂无评论

Bioactive Materials 2021年第10期6卷 3424-3436页

作者： Maria Wątroba Wiktor Bednarczyk Jakub Kawałko Piotr Bała AGH University of Science and Technology Faculty of Metals Engineering and Industrial Computer ScienceAl.A.Mickiewicza 3030-059 KrakowPoland AGH University of Science and Technology Academic Centre for Materials and NanotechnologyAl.A.Mickiewicza 3030-059 KrakowPoland

In recent years,Zn-based materials have been extensively investigated as potential candidates for biodegradable implant *** introduction of alloying elements providing solid-solution strengthening and second phase strengthening seems crucial to provide a suitable platform for the thermo-mechanical strengthening of Zn *** this study,a systematic investigation of the microstructure,crystallographic texture,phase composition,and mechanical properties of a Zn-3Ag-0.5Mg(wt%)alloy processed through combined hot extrusion(HE)and cold rolling(CR),followed by short-time heat treatment(CR+HT)at 200◦C was ***,the influence of different annealing temperatures on the microstructure and mechanical properties was *** adequate combination of processing conditions during CR and HT successfully addressed brittleness obtained in the high-strength HE Zn-3Ag-0.5Mg *** controlling the microstructure,the most promising results were obtained in the sample subjected to 50%CR reduction and 5-min annealing,which were:ultimate tensile strength of 432 MPa,yield strength of 385 MPa,total elongation to failure of 34%,and Vickers microhardness of 125 *** obtained properties clearly exceed the mechanical benchmarks for biodegradable implant *** on the conducted investigation,brittle multi-phase Zn alloys’mechanical performance can be substantially enhanced to provide sufficient plasticity by grain refinement through cold deformation process,followed by short-time annealing to restore proper strength.

关键词： Zinc alloys Cold rolling Heat treatment Mechanical properties Grain refinement

Fast Click-Through Rate Estimation Using Data Aggregates 21st

学校读者我要写书评

暂无评论

Fast Click-Through Rate Estimation Using Data Aggregates

21st International Conference on Computational science, ICCS 2021

作者： Wiatr, Roman Slota, Renata G. Kitowski, Jacek Faculty of Computer Science Electronics and Telecommunication Institute of Computer Science AGH University of Science and Technology Mickiewicza 30 Krakow30-059 Poland Academic Computer Centre CYFRONET AGH AGH University of Science and Technology Nawojki 11 Krakow30-950 Poland

ISBN: (纸本)9783030779603

Click-Through Rate estimation is a crucial prediction task in Real-Time Bidding environments prevalent in display advertising. The estimation provides information on how to trade user visits in various systems. Logistic Regression is a popular choice as the model for this task. Due to the amount, dimensionality and sparsity of data, it is challenging to train and evaluate the model. One of the techniques to reduce the training and evaluation cost is dimensionality reduction. In this work, we present Aggregate Encoding, a technique for dimensionality reduction using data aggregates. Our approach is to build aggregate-based estimators and use them as an ensemble of models weighted by logistic regression. The novelty of our work is the separation of feature values according to the value frequency, to better utilise regularization. For our experiments, we use the iPinYou data set, but this approach is universal and can be applied to other problems requiring dimensionality reduction of sparse categorical data. © 2021, Springer Nature Switzerland AG.

关键词： Aggregates

Unveiling the Power of Sparse Neural Networks for Feature Selection

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Atashgahi, Zahra Liu, Tennison Pechenizkiy, Mykola Veldhuis, Raymond Mocanu, Decebal Constantin van der Schaar, Mihaela Faculty of Electrical Engineering Mathematics and Computer Science University of Twente Netherlands Department of Applied Mathematics and Theoretical Physics University of Cambridge United Kingdom Department of Mathematics and Computer Science Eindhoven University of Technology Netherlands Department of Computer Science University of Luxembourg Luxembourg

Sparse Neural Networks (SNNs) have emerged as powerful tools for efficient feature selection. Leveraging the dynamic sparse training (DST) algorithms within SNNs has demonstrated promising feature selection capabilities while drastically reducing computational overheads. Despite these advancements, several critical aspects remain insufficiently explored for feature selection. Questions persist regarding the choice of the DST algorithm for network training, the choice of metric for ranking features/neurons, and the comparative performance of these methods across diverse datasets when compared to dense networks. This paper addresses these gaps by presenting a comprehensive systematic analysis of feature selection with sparse neural networks. Moreover, we introduce a novel metric considering sparse neural network characteristics, which is designed to quantify feature importance within the context of SNNs. Our findings show that feature selection with SNNs trained with DST algorithms can achieve, on average, more than 50% memory and 55% FLOPs reduction compared to the dense networks, while outperforming them in terms of the quality of the selected features. Our code and the supplementary material are available on GitHub (https://***/zahraatashgahi/Neuron-Attribution). © 2024, CC BY-SA.

关键词： Feature Selection