the University of Computer Science is an educational-production centre that develops computer applications and services in different areas of knowledge. Because of the added value that the developed applications get w...
详细信息
ISBN:
(纸本)9783030525750;9783030525743
the University of Computer Science is an educational-production centre that develops computer applications and services in different areas of knowledge. Because of the added value that the developed applications get when Artificial Intelligence techniques are used for decision making, the academic formation of the students in the different disciplines of the profession with emphasis on Artificial Intelligence is of high interest. this work describes a computational tool to support teaching that uses the conceptual algorithms of logical combinatorial recognition to determine the architecture and the set of initial weights of an artificial neural network, which contributes to the temporal efficiency of the learning process of the network and the efficiency of the classification. Experiments using Friedman's test and cross validation method demonstrate the applicability of this hybrid model in a Multilayer Perceptron.
the proceedings contain 99 papers. the topics discussed include: command filter based event-triggered adaptive control for strict-feedback nonlinear systems with full state constraints;PointDet: An object detection fr...
ISBN:
(纸本)9781665412667
the proceedings contain 99 papers. the topics discussed include: command filter based event-triggered adaptive control for strict-feedback nonlinear systems with full state constraints;PointDet: An object detection framework based on human local features in the task of identifying violations;FPGA implementation of object detection accelerator based on Vitis-AI;digital twin enhanced assembly based on deep reinforcement learning;demand forecasting for shared umbrella using BP neural network;fully circuit implementation of a two-layer memristive neural network for patternrecognition;low-rank transfer learning for multi-stream data classification;using deep learning techniques to predict 10- year us treasury yield;and SegPoseNet: segmentation-guided 3D hand pose estimation.
the main problem solved by the authors is to reduce the complexity of creating and maintaining systems with knowledge bases and increase their viability. the authors see its solution in the classification of intellect...
详细信息
Place recognition is an important topic in many vision-based applications. For this purpose, many traditional and artificial intelligence-based methods have been proposed. Despite years of expertise accrued in this ar...
详细信息
When people come together as part of a community, oriented towards a collective activity, over an extended period of time, they develop and maintain different routines for the way they are organized, delegate, and car...
详细信息
As is known to all, withthe continuous development of computer technology, computer viruses are also increasingly prominent, especially the phenomenon of computer virus database is increasingly rampant, and the emerg...
详细信息
Convolutional Neural Network (CNN) is widely acknowledged as an effective machine learning model for various detection and recognition tasks. However, CNN often requires a significant amount of hardware resources and ...
详细信息
ISBN:
(数字)9783031168185
ISBN:
(纸本)9783031168185;9783031168178
Convolutional Neural Network (CNN) is widely acknowledged as an effective machine learning model for various detection and recognition tasks. However, CNN often requires a significant amount of hardware resources and is high in its power consumption. this hinders the widespread deployment of CNN model in embedded systems and wearable devices. therefore, stochastic computing (SC) which leverages the power-accuracy trade-off, began to gain popularity in various neural network (NN) implementations. this paper presents an improved SC multiply-and-accumulate (MAC) unit that can be utilized as convolution engines in CNN. the proposed SC-MAC is operated using deterministic sequence and the design achieves latency and power reductions through parallelism and split mechanism optimizations. Furthermore, we also introduce decoder-based Stochastic Number Generator (SNG) that is capable of generating uncorrelated and segmented stochastic number (SN) without using random sources. the proposed deterministic and split SC-MAC is synthesized using typical libraries of UMC 40 nm technology for detailed hardware evaluation. the functionality of the presented SC-MAC is also verified in CNN using the MNIST dataset. Overall, our SC-MAC is proven to achieve higher power efficiency (GMACS/mW) and lower in energy consumption (pJ/MAC) as compared to the related works.
the term Deep Learning can be termed as the subset of artificial intelligence with multiple network layers forming neural patterns. People’s interest in having the knowledge of deep hidden layers have recently booste...
详细信息
As quantum computers are becoming real, they have the inherent potential to significantly impact many application domains. In this paper we outline the fundamentals about programming quantum computers and show that qu...
详细信息
ISBN:
(数字)9783030723699
ISBN:
(纸本)9783030723682;9783030723699
As quantum computers are becoming real, they have the inherent potential to significantly impact many application domains. In this paper we outline the fundamentals about programming quantum computers and show that quantum programs are typically hybrid consisting of a mixture of classical parts and quantum parts. Withthe advent of quantum computers in the cloud, the cloud is a fine environment for performing quantum programs. the tool chain available for creating and running such programs is sketched. As an exemplary problem we discuss efforts to implement quantum programs that are hardware independent. A use case from quantum humanities is discussed, hinting which applications in this domain can already be used in the field of (quantum) machine learning. Finally, a collaborative platform for solving problems with quantum computers - that is currently under construction - is presented.
the proceedings contain 25 papers. the special focus in this conference is on Numerical Geometry, Grid Generation, and Scientific computing. the topics include: Efficient Steady Flow Computations with Exponential Mult...
ISBN:
(纸本)9783030767976
the proceedings contain 25 papers. the special focus in this conference is on Numerical Geometry, Grid Generation, and Scientific computing. the topics include: Efficient Steady Flow Computations with Exponential Multigrid Methods;exponential Time Integrators for Unsteady Advection–Diffusion Problems on Refined Meshes;a Proof of the Invariant-Based Formula for the Linking Number and Its Asymptotic Behaviour;the Singularity Set of Optimal Transportation Maps;polygonal and Polyhedral Delaunay Meshing;on Decomposition of Embedded Prismatoids in 3 Without Additional Points;out-of-core Constrained Delaunay Tetrahedralizations for Large Scenes;size Gradation Control for Anisotropic Hybrid Meshes;Adjoint Computation on Anisotropic Meshes in High-fidelity RANS Simulations;preface;foreword;local Groups in Delone Sets;moving Deforming Mesh Generation Based on the Quasi-Isometric Functional;Adaptive Grids for Non-monotone Waves and Instabilities in a Non-equilibrium PDE Model;RBF-VerBSS Hybrid Method for Mesh Deformation;a Uniform Convergence Analysis for a Bakhvalov-Type Mesh with an Explicitly Defined Transition Point;on a Comprehensive Grid for Solving Problems Having Exponential or Power-of-First-Type Layers;preserved Structure Constants for Red Refinements of Product Elements;global Parametrization Based on Ginzburg-Landau Functional;parametrization of Plane Irregular Regions: A Semi-automatic Approach I;a Hybrid Approach to Fast Indirect Quadrilateral Mesh Generation;hexahedral Mesh Generation Using Voxel Field Recovery;manifolds of Triangulations, Braid Groups of Manifolds, and the Groups Γnk;generation of Boundary Layer Meshes by the Enhanced Jump-and-Walk Method with a Fast Collision Detecting Algorithm.
暂无评论