LDPCD: A Novel Method for Locally Differentially Private Community Detection

As one of the cores of data analysis in large social networks, community detection has become a hot research topic in recent years. However, user’s real social relationship may be at risk of privacy leakage and threatened by inference attacks because of the semitrusted server. As a result, community detection in social graphs under local differential privacy has gradually aroused the interest of industry and academia. On the one hand, the distortion of user’s real data caused by existing privacy-preserving mechanisms can have a serious impact on the mining process of densely connected local graph structure, resulting in low utility of the final community division. On the other hand, private community detection requires to use the results of multiple user-server interactions to adjust user’s partition, which inevitably leads to excessive allocation of privacy budget and large error of perturbed data. For these reasons, a new community detection method based on the local differential privacy model (named LDPCD) is proposed in this paper. Due to the introduction of truncated Laplace mechanism, the accuracy of user perturbation data is improved. In addition, the community divisive algorithm based on extremal optimization (EO) is also refined to reduce the number of interactions between users and the server. Thus, the total privacy overhead is reduced and strong privacy protection is guaranteed. Finally, LDPCD is applied in two commonly used real-world datasets, and its advantage is experimentally validated compared with two state-of-the-art methods.

On Quantization Errors in Approximate and Sample Entropy

Approximate and sample entropies are acclaimed tools for quantifying the regularity and unpredictability of time series. This paper analyses the causes of their inconsistencies. It is shown that the major problem is a coarse quantization of matching probabilities, causing a large error between their estimated and true values. Error distribution is symmetric, so in sample entropy, where matching probabilities are directly summed, errors cancel each other. In approximate entropy, errors are accumulating, as sums involve logarithms of matching probabilities. Increasing the time series length increases the number of quantization levels, and errors in entropy disappear both in approximate and in sample entropies. The distribution of time series also affects the errors. If it is asymmetric, the matching probabilities are asymmetric as well, so the matching probability errors cease to be mutually canceled and cause a persistent entropy error. Despite the accepted opinion, the influence of self-matching is marginal as it just shifts the error distribution along the error axis by the matching probability quant. Artificial lengthening the time series by interpolation, on the other hand, induces large error as interpolated samples are statistically dependent and destroy the level of unpredictability that is inherent to the original signal.

Eliminating the Influence of Axle Parameters in Influence Line Identification

Accurate and rapid acquisition of the strain influence line of continuous beam plays a positive role in promoting the wide application of structural health monitoring. The structural response obtained from the sensors is used to estimate the strain influence line. However, most estimation methods ignore the influence of axle parameters on the structural response, resulting in a large error in identifying the strain influence line. This paper presents a method for eliminating the influence of axle parameters of moving vehicles on strain responses to estimate the strain influence line of continuous beams based on the long-gauge strain sensing technology. By analysing the mechanical characteristics of the multi-span continuous beam, a theoretical strain influence line expression is first established to obtain the strain influence line of the continuous beam accurately. The structural response only caused by axle weight, obtained by eliminating the influence of axle parameters, is then estimated for calibrating the theoretical strain influence line. Finally, different lane tests are also considered to solve the influence of different transverse position relations on the proposed method between the monitoring unit and the lane. Finally, numerical simulations are adopted to illustrate the effectiveness of the proposed identification method by simulating the strain time histories induced by a multi-axle vehicle. A field test also demonstrates the validity and feasibility of this method.

Automatic verification method of progressive image model matching information based on machine learning

In order to solve the problems of large error and low performance of traditional progressive image model matching information checking methods, an automatic progressive image model matching information checking method based on machine learning is proposed. The generation method of progressive image is analyzed, and the target image sample is obtained. On this basis, machine learning algorithm is used to segment progressive image samples. In each image segmentation part, crawler technology is used to automatically collect progressive image model matching information, and under the constraint of image model matching information checking standard, automatic checking of progressive image model matching information is realized from geometric structure, image content and other aspects. Experimental results show that the verification error of the design method is reduced by 0.687 Mb, and the quality of progressive image is improved.

Pacing Electrocardiogram Detection With Memory-Based Autoencoder and Metric Learning

Remote ECG diagnosis has been widely used in the clinical ECG workflow. Especially for patients with pacemaker, in the limited information of patient's medical history, doctors need to determine whether the patient is wearing a pacemaker and also diagnose other abnormalities. An automatic detection pacing ECG method can help cardiologists reduce the workload and the rates of misdiagnosis. In this paper, we propose a novel autoencoder framework that can detect the pacing ECG from the remote ECG. First, we design a memory module in the traditional autoencoder. The memory module is to record and query the typical features of the training pacing ECG type. The framework does not directly feed features of the encoder into the decoder but uses the features to retrieve the most relevant items in the memory module. In the training process, the memory items are updated to represent the latent features of the input pacing ECG. In the detection process, the reconstruction data of the decoder is obtained by the fusion features in the memory module. Therefore, the reconstructed data of the decoder tends to be close to the pacing ECG. Meanwhile, we introduce an objective function based on the idea of metric learning. In the context of pacing ECG detection, comparing the error of objective function of the input data and reconstructed data can be used as an indicator of detection. According to the objective function, if the input data does not belong to pacing ECG, the objective function may get a large error. Furthermore, we introduce a new database named the pacing ECG database including 800 patients with a total of 8,000 heartbeats. Experimental results demonstrate that our method achieves an average F1-score of 0.918. To further validate the generalization of the proposed method, we also experiment on a widely used MIT-BIH arrhythmia database.

SMAP Salinity Retrievals near the Sea-Ice Edge Using Multi-Channel AMSR2 Brightness Temperatures

Sea-ice contamination in the antenna field of view constitutes a large error source in retrieving sea-surface salinity (SSS) with the spaceborne Soil Moisture Active Passive (SMAP) L-band radiometer. This is a major obstacle in the current NASA/Remote Sensing Systems (RSS) SMAP SSS retrieval algorithm in regards to obtaining accurate SSS measurements in the polar oceans. Our analysis finds a strong correlation between 8-day averaged SMAP L-band brightness temperature (TB) bias and TB measurements from the Advanced Microwave Scanning Radiometer (AMSR2) in the C-through Ka-band frequency range for sea-ice contaminated ocean scenes. We show how this correlation can be employed to develop: (1) a discriminant analysis that is able to reliably flag the SMAP observations for sea-ice contamination and (2) subsequently remove the sea-ice contamination from the SMAP observations, which results in significantly more accurate SMAP SSS retrievals near the sea-ice edge. We provide a case study that evaluates the performance of the proposed sea-ice flagging and correction algorithm. Our method is also able to detect drifting icebergs, which go often undetected in many available standard sea-ice products and thus result in spurious SMAP SSS retrievals.

Mobile Industrial Robots Localization Algorithm Based on Improved Multidimensional Scale and Received Signal Strength Indication

Aiming at the problem of large error in the location algorithm based on MDS-MAP when the distance between mobile industrial robots is not measurable, a mobile industrial robot location algorithm based on improved MDS-MAP is proposed. Experimental simulation shows that the algorithm can achieve good positioning effect. When the distance between mobile industrial robots is measurable, the positioning algorithm based on RSSI achieves good positioning effect. Therefore, this paper discusses the influence of different anchor robot selection methods on the positioning accuracy of RSSI positioning algorithm. The experimental simulation shows that when the selection method of anchoring robot is that the unknown robot with adjacent anchoring robot uses the original anchoring robot for positioning and the unknown robot without anchoring robot uses the adjacent positioning robot as the anchoring robot for positioning, its positioning effect is the best, and it can still achieve good positioning effect when there are few anchoring robots.

Modeling, System Measurements and Controller Investigation of a Small Battery-Powered Fixed-Wing UAV

In this paper, a complete set of nonlinear modeling and controller design process for a small electric fixed-wing unmanned aerial vehicle (UAV) is presented. The nonlinear mathematical model and aerodynamic model of the small fixed-wing UAV are derived. The computational fluid dynamics (CFD) method was used to obtain the aerodynamic coefficients of the UAV, and the models of propulsion system components were established through experiments. Since the linearized and decoupled model of the fixed-wing UAV has a large error, a nonlinear model is established based on Simulink, which is utilized to design and verify the control algorithms. Based on the established nonlinear model, a stability controller, path following controller and path management controller of the aircraft are set up. The results indicate that system parameters of the aircraft can be quickly acquired and an efficient and practical model can be established by the methods. In addition, the controller designed and applied in this paper has good performance and small steady-state error, which can meet the basic flight mission requirements, including stability of flight attitude, path following and switching of different waypoints. These modeling and control methods can also be employed in other small battery-powered fixed-wing UAV projects.

Research on Efficient Reinforcement Learning for Adaptive Frequency-Agility Radar

Modern radar jamming scenarios are complex and changeable. In order to improve the adaptability of frequency-agile radar under complex environmental conditions, reinforcement learning (RL) is introduced into the radar anti-jamming research. There are two aspects of the radar system that do not obey with the Markov decision process (MDP), which is the basic theory of RL: Firstly, the radar cannot confirm the interference rules of the jammer in advance, resulting in unclear environmental boundaries; secondly, the radar has frequency-agility characteristics, which does not meet the sequence change requirements of the MDP. As the existing RL algorithm is directly applied to the radar system, there would be problems, such as low sample utilization rate, poor computational efficiency and large error oscillation amplitude. In this paper, an adaptive frequency agile radar anti-jamming efficient RL model is proposed. First, a radar-jammer system model based on Markov game (MG) established, and the Nash equilibrium point determined and set as a dynamic environment boundary. Subsequently, the state and behavioral structure of RL model is improved to be suitable for processing frequency-agile data. Experiments that our proposal effectively the anti-jamming performance and efficiency of frequency-agile radar.

Automation of Collection of Primary Data for Development of a Passenger Origin-Destination Trip Correspondence Matrix Based on Computer Vision and Neural Network Technologies

The origin-destination trip matrix is a fundamental characteristic of a transport network, and development of a reliable correspondence matrix is the most important task in organising passenger traffic. It is the basis on which the public transport route network of a city (region) is built and optimised.Currently, collection of initial information for construction of a travel correspondence matrix is carried out through field surveys comprising questionnaire surveys of the population; accounting for movement of passengers according to the coupons issued to them; checkers, tellers manually counting passengers in vehicle compartments; simple surveys of passengers. Besides, mathematical modelling is used based on statistical data on the number of residents in various districts of the city, employees in enterprises and students in educational institutions, as well as on available data on the characteristics of passenger traffic along certain routes. All these surveys are very expensive and are carried out once over few years; they give a large error, which is why decisions made on the basis of these data are far from being optimal.There are a lot of solutions in the software and hardware market that provide automated collection of data on passenger flows. They are based on the use of infrared sensors or of video recording. However, none of these systems provide information about the points of entry and exit of each passenger. The objective of this study was to develop methods for automating the collection of reliable information about passenger trips, that will be the base for building up-to-date and reliable passenger trip correspondence matrices. This task can be solved by constant monitoring of passengers’ trips with fixing places of entry and exit of each passenger.The study describes the possibility of creating software based on computer vision and artificial intelligence which will provide automation of collection of primary information about travel of each passenger from the place of boarding into the vehicle to exit from it, that is, automation of data generation to build a passenger trip correspondence matrix.