Abstract Objective Compared with other regions in the world, transmission characteristics of COVID-19 epidemic Africa are more obvious, has a unique mode this region; At same time, data related to is characterized by low quality and incomplete coverage, which makes prediction method suitable for unable achieve good results Africa. In order solve above problems, paper proposes that nests in-depth learning mechanism model. From experimental results, it can better problems adapt African...

This article examines the influence of perceived personal impact on third-person perception and protective behaviors in connection with exposure to media coverage H1N1 swine flu pandemic. Survey results show that disease spreading local community positively predicts effects self others. However, its self-evaluation effect is more salient and, thus, negatively influences perception. In terms corresponding behaviors, people's concerns for their own safety rather than others' predict intention...

Machine learning-base fire detection methods play a vital role in the current forest monitoring systems. In literature, is usually viewed as two-class (or multi-class) classification task. However, this violates most foundational hypothesis machine learning, e.g., independent and identical distribution (i.i.d.), especially for non-fire samples drawn from complex background. Moreover, omni-directional video -monitoring, background also always changing, which leads violation to worse...

This study focuses on enhancing the precision of epidemic time series data prediction by integrating Gated Recurrent Unit (GRU) into a Graph Neural Network (GNN), forming GRGNN. The accuracy GNN (Graph Network) network with introduced GRU (Gated Units) is validated comparing it seven commonly used methods. GRGNN methodology involves multivariate using improved integration Units). Additionally, Graphical Fourier Transform (GFT) and Discrete (DFT) are introduced. GFT captures inter-sequence...

This paper takes on the central heating secondary network and establishes a two-level diagnosis model of leakage fault pipe based deep belief (DBN) under condition constant small supply flow quality regulation. Firstly, hydraulic calculation is established with graph theory, which provides pressure changes monitoring points in network. Then, first-level diagnostic for designed to diagnose leaky segment by using Based results model, each treated as unit second-level then developed predict...

In the Internet of Things (IoT) scenario, hardware resources in edge computing paradigm are limited, so applications cloud not suitable for computing. New system architecture and key technologies needed application implement deployment Based on performance constraints scene functional requirements under computing, this paper designs develops a microservice So it introduces design And analyzes system. It selects PostgreSQL database to achieve data storage, Flask framework build back-end...

This paper proposes a prediction approach on energy consumption for public buildings based mind evolutionary algorithm and BP neural network. The actual real-time data of some layer in building can be obtained online by our implemented monitoring system, then several key factors which affect analyzed determined correlation analysis method. By using the algorithm, ideal weight values threshold network are calculated, solve its problems low efficiency slow convergence. Finally, performance...

Aiming at the large amount of data and complex structure heating pipe network system, this paper studies fault diagnosis method leakage based on deep belief network. Firstly, using theory graph to establish hydraulic calculation model condition network, then according pressure change value monitoring point in network; is used leakage, feasibility verified by experiments. The experimental results show that proposed has higher accuracy for prediction segments, superior BP (Back Propagation...

Anomalies in vaccine data are often encountered when projecting the active cases Africa. This issue could severely impact accuracy of expected case, thereby affecting future government intervention protection policies etc. In this paper, a new method to accurately predict considering anomalies is presented. Firstly, an autoencoder used detect anomalies, Then, order impute missing values complete African vaccination small-scale k nearest neighbor (ssKNN) with upper and lower bounds developed....

Tunnel LED lighting system is an important infrastructure to ensure the safety and comfort of driving in highway tunnels. At present, traditional method for tunnel patrol missing-lamp (or faulty-lamp) condition carried out with manual labor, which has high labor cost low efficiency. In this paper, automatic robust detection missing lamps tunnels based on vehicle-mounted video processing proposed. This includes following three aspects. Firstly, a three-phase algorithm was proposed detect LEDs...

In the merge areas of highway, vehicles often misjudge traffic situations themselves and their adjacent lanes, so it is easy to cause problems such as sudden stops or collisions, which can lead jams. Advances in Connected Automated Vehicles allow us build a computational framework for real-time control coordination vehicles. This paper establishes decentralized optimal solve congestion two areas, reduce travel time fuel consumption. Firstly, we determine priority on different roads...

In the face of rapid development IoT, traditional cloud computing, while providing services for mobile intelligent devices, also brings problems such as network bandwidth occupation and communication delay. Edge computing can solve these to some extent. But too many offloaded tasks may result in an edge server becoming overloaded execution delay increasing, resources idle devices are not fully utilized. order make full use local reduce tasks, this paper takes average time all end...

Tunnel LED lighting system is an important infrastructure to ensure the safety and comfort of driving in highway tunnels. There had been a series researches about analyzing these lamps working state by detection missing In this paper, automatic tracking method for salient line segments tunnels, left right critical boundaries between top dark area tunnel white side wall, proposed vehicle-mounted video processing. This includes following two aspects. Firstly, three-phase algorithm was detect...

NA

The Robotics Wushu Competition is a robotics competition that integrates intelligent technology with traditional Chinese culture, maker spirit, and ideas. It highly confrontational ornamental project in competitions. This paper discusses the current situation, development status, settings of three major national competitions, focusing on interpretation several projects, such as humanoid autonomous fighting, body-feeling wheeled visual confrontation, unmanned aerial vehicle (UAV)...

In order to solve the problem that it is difficult efficiently identify homologous complaints and reports in dealing with reports, this paper proposes detect complaint decomposed attention network. Firstly, TextCNN used extract local features of text, then matching task between carried out. Due high noise method a few key better than capturing global text when judging whether two reporting tasks are same origin events. The resolvable network same-origin into recognition channels, judged...