A5010106771- Network Security and Intrusion Detection
- Financial Distress and Bankruptcy Prediction
- Advanced Battery Technologies Research
- Customer churn and segmentation
- Anomaly Detection Techniques and Applications
- Internet Traffic Analysis and Secure E-voting
- Robotic Path Planning Algorithms
- Advanced Computational Techniques and Applications
- Interstitial Lung Diseases and Idiopathic Pulmonary Fibrosis
- Data Mining Algorithms and Applications
- Advanced Algorithms and Applications
- Chaos-based Image/Signal Encryption
- Medical Imaging and Pathology Studies
- Guidance and Control Systems
- Pulmonary Hypertension Research and Treatments
- Adaptive Control of Nonlinear Systems
- Advanced Malware Detection Techniques
- Flowering Plant Growth and Cultivation
- Electric and Hybrid Vehicle Technologies
- Numerical methods for differential equations
- Energy and Environment Impacts
- IoT-based Smart Home Systems
- Data Management and Algorithms
- Aerospace Engineering and Control Systems
- Plant Taxonomy and Phylogenetics
Beijing Jingshida Electromechanical Equipment Research Institute
2019-2024
Taiyuan University of Science and Technology
2024
North China University of Science and Technology
2011-2023
Computer Emergency Response Team
2019
Zhejiang University
2003-2011
Skin disease is one of the most common diseases. Due to intricate categories skin diseases, their symptoms being very similar in early stage, and lesion samples extremely unbalanced, classification challenging. At same time, under conditions limited data, generalization ability a single reliable convolutional neural network model weak, feature extraction insufficient, accuracy low. Therefore, this paper, we proposed for based on fusion. Through fusion, deep shallow introduction an attention...
Dijkstra algorithm is a classic of computing the shortest path between two points. In past amount urban vehicles small, so optimal one. However, with increasing cars, traffic congestion serious. The calculated by using traditional still path, but it not necessarily optimal. Therefore, in order to solve route choice problem improved put forward, which introduces weight function. Experiments show that can better meet needs drivers on path.
With the rapid development of LAN, security LAN has been becoming serious, especially malicious network attack is one vital problems. In this paper, with web log mining technique, files which are generated in server analysed by outlier mining, can effectively supervise and promise operation techniques.
Successive Over-Relaxation Iterative Algorithm (hereinafter referred to as SOR Algorithm) is one of the effective methods for solving large-scale sparse matrix equation set, and first order linear stationary iterative method. Starting with introduction system algebraic equations, this paper discusses astringency judgement conditions importance selection convergence factor, provides MATLAB program Algorithm, equations; algorithm; symmetric positive definite matrices.
To improve the fuel economy and battery life of extended‐range electric vehicles (EREVs), an energy management strategy (EMS) based on bargaining game (BG) is proposed. This optimizes overall by optimizing power allocation. First, a model for EREVs developed. The considers auxiliary unit (APU) as two parties game. allocation between APU both consumption life. And it also improves benefit vehicle. Second, effect discount factor control results analyzed to obtain best optimization results....
This paper presented that intrusion detection is an effective technique to protect Web server. Misused-based used detect Web-based attacks. It went deeply into Decision Tree technique; the traditional misused-based matching method was improved by using decision tree and speed of process has been significantly improved. Its efficiency verified example.
Our team has won the Grand Champion (Tie) of PAKDD-2007 data mining competition. The task is to score credit card customers a consumer finance company according likelihood that take up home loans offered by company. This report presents our solution for this business problem. TreeNet and logistic regression are algorithms used in project. final based on cross-algorithm ensemble two within-algorithm ensembles regression. Finally, some discussions from presented.
Lyapunov index is one of criteria for testing whether the system in a chaotic state, and its value represents developed level state. To study characteristic cascade reveal disturbance mechanism among subsystems system, following researches are carried out. First, model constructed from viewpoint pseudo noise disturbance, difference between without with external influence investigated. Then conclusion that disturbances can be considered as drawn. Second, proved not algebraic sum each...
Meeting distributed detection's requirements of accuracy, expansibility and portability, ontology is introduced to organize systematize all intrusion detection behavior at any level, consequently reducing ambiguity supporting share. The paper presented ontology-based model with advantages hierarchical cooperative models, then the deduction expert system was used detect complex attacks.
In fractional Fourier transform (FRFT) image encryption algorithm, according to the problem that one plain-text is correlated with cipher-text encrypted, which can not resist chosen plaintext and attack, double-feedback method based on chaotic dynamic grouping oriented FRFT was proposed. Two systems were adopted in this method. Firstly, feedback introduced drive two systems. And driven by utilized encrypt: keys of system 1 realize plain later FRFT, first obtained. Then 2 obtain second...
Aiming at the web attacks, this paper presented a program-based intrusion detection system to protect Web server based on analysis of data source and technique. Then, it extended object safety specification so as meet requirements server. In conclusion, new technology, can be applied detect attacks against vulnerabilities effectively.
In this paper, influences of actuator nonlinearities and their suppression methods are studied. Firstly, the dead zone backlash characteristics on control performance statically unstable missiles analyzed in detail. Then based cascade method active disturbance rejection (ADRC) theory, a nonlinear "double-layer compensation" for is proposed. Finally, simulation verification carried out. Simulation results show that ADRC approach can significantly suppress adverse effects such as system...