We study the problem of business model mining and prediction in the e-commerce context. Unlike most existing approaches where this is typically formulated as a regression problem or a time-series prediction problem, we take a different formulation to this problem by noting that these existing approaches fail to consider the potential relationships both among the consumers (consumer influence) and among the shops (competitions or collaborations). Taking this observation into consideration, we propose a new method for e-commerce business model mining and prediction, called EBMM, which combines regression with community analysis. The challenge is that the links in the network are typically not directly observed, which is addressed by applying information diffusion theory through the consumer-shop network. Extensive evaluations using Alibaba Group e-commerce data demonstrate the promise and superiority of EBMM to the state-of-the-art methods in terms of business model mining and prediction.
Recently, privacy concerns about data collection have received an increasing amount of attention. In data collection process, a data collector (an agency) assumed that all respondents would be comfortable with submitting their data if the published data was anonymous. We believe that this assumption is not realistic because the increase in privacy concerns causes some respondents to refuse participation or to submit inaccurate data to such agencies. If respondents submit inaccurate data, then the usefulness of the results from analysis of the collected data cannot be guaranteed. Furthermore, we note that the level of anonymity (i.e., k-anonymity) guaranteed by an agency cannot be verified by respondents since they generally do not have access to all of the data that is released. Therefore, we introduce the notion of ki-anonymity, where ki is the level of anonymity preferred by each respondent i. Instead of placing full trust in an agency, our solution increases respondent confidence by allowing each to decide the preferred level of protection. As such, our protocol ensures that respondents achieve their preferred ki-anonymity during data collection and guarantees that the collected records are genuine and useful for data analysis.
End-to-end delay measurement has been an essential element in the deployment of real-time services in networked systems. Traditional methods of delay measurement based on time domain analysis, however, are not efficient as the network scale and the complexity increase. We propose a novel theoretical framework to analyze the end-to-end delay distributions of networked systems from the frequency domain. We use a signal flow graph to model the delay distribution of a networked system and prove that the end-to-end delay distribution is indeed the inverse Laplace transform of the transfer function of the signal flow graph. Two efficient methods, Cramer’s rule-based method and the Mason gain rule-based method, are adopted to obtain the transfer function. By analyzing the time responses of the transfer function, we obtain the end-to-end delay distribution. Based on our framework, we propose an efficient method using the dominant poles of the transfer function to work out the bottleneck links of the network. Moreover, we use the framework to study the network protocol performance. Theoretical analysis and extensive evaluations show the effectiveness of the proposed approach.
The symbolic representation of time series has attracted much research interest recently. The high dimensionality typical of the data is challenging, especially as the time series becomes longer. The wide distribution of sensors collecting more and more data exacerbates the problem. Representing a time series effectively is an essential task for decision-making activities such as classification, prediction, and knowledge discovery. In this paper, we propose a new symbolic representation method for long time series based on trend features, called trend feature symbolic approximation (TFSA). The method uses a two-step mechanism to segment long time series rapidly. Unlike some previous symbolic methods, it focuses on retaining most of the trend features and patterns of the original series. A time series is represented by trend symbols, which are also suitable for use in knowledge discovery, such as association rules mining. TFSA provides the lower bounding guarantee. Experimental results show that, compared with some previous methods, it not only has better segmentation efficiency and classification accuracy, but also is applicable for use in knowledge discovery from time series.
This paper investigates the H∞ trajectory tracking control for a class of nonlinear systems with timevarying delays by virtue of Lyapunov-Krasovskii stability theory and the linear matrix inequality (LMI) technique. A unified model consisting of a linear delayed dynamic system and a bounded static nonlinear operator is introduced, which covers most of the nonlinear systems with bounded nonlinear terms, such as the one-link robotic manipulator, chaotic systems, complex networks, the continuous stirred tank reactor (CSTR), and the standard genetic regulatory network (SGRN). First, the definition of the tracking control is given. Second, the H∞ performance analysis of the closed-loop system including this unified model, reference model, and state feedback controller is presented. Then criteria on the tracking controller design are derived in terms of LMIs such that the output of the closed-loop system tracks the given reference signal in the H∞ sense. The reference model adopted here is modified to be more flexible. A scaling factor is introduced to deal with the disturbance such that the control precision is improved. Finally, a CSTR system is provided to demonstrate the effectiveness of the established control laws.
We propose a novel series transformer based diode-bridge-type solid state fault current limiter (SSFCL). To control the fault current, a series RLC branch is connected to the secondary side of an isolation series transformer. Based on this RLC branch, two current limiting modes are created. In the first mode, R and C are bypassed via a paralleled power electronic switch (insulated-gate bipolar transistor, IGBT) and L remains connected to the secondary side of the transformer as a DC reactor. In the second mode, the series reactor impedance is not enough to limit the fault current. In this case, the fault current can be controlled by selecting a proper on-off duration of the parallel IGBT, across the series damping resistor (R) and capacitor, which inserts high impedance into the line to limit the fault current. Then, by controlling the magnitude of the DC reactor current, the fault current is reduced and the voltage of the point of common coupling (PCC) is kept at an acceptable level. In addition, in the new SSFCL, the series RC branch, connected in parallel with the IGBT, serves as a snubber circuit for decreasing the transient recovery voltage (TRV) of the IGBT during on-off states. Therefore, the power quality indices can be improved. The measurement results of a built prototype are presented to support the simulation and theoretical studies. The proposed SSFCL can limit the fault current without any delay and successfully smooth the fault current waveform.
A cepstrum moving target detection (CEPMTD) algorithm based on cepstrum techniques is proposed for passive coherent location (PCL) radar systems. The primary cepstrum techniques are of great success in recognizing the arrival times of static target echoes. To estimate the Doppler frequencies of moving targets, we divide the radar data into a large number of segments, and reformat these segments into a detection matrix. Applying the cepstrum and the Fourier transform to the fast and slow time dimensions respectively, we can obtain the range information and Doppler information of the moving targets. Based on the CEPMTD outlined above, an improved CEPMTD algorithm is proposed to improve the detection performance. Theoretical analyses show that only the target’s peak can be coherently added. The performance of the improved CEPMTD is initially validated by simulations, and then by experiments. The simulation results show that the detection performance of the improved CEPMTD algorithm is 13.3 dB better than that of the CEPMTD algorithm and 6.4 dB better than that of the classical detection algorithm based on the radar cross ambiguity function (CAF). The experiment results show that the detection performance of the improved CEPMTD algorithm is 1.63 dB better than that of the radar CAF.
Weak L1 signal acquisition in a high dynamic environment primarily faces a challenge: the integration peak is negatively influenced by the possible bit sign reversal every 20 ms and the frequency error. The block accumulating semi-coherent integration of correlations (BASIC) is a state-of-the-art method, but calculating the inter-block conjugate products restricts BASIC in a low signal-to-noise ratio (SNR) acquisition. We propose a block zero-padding method based on a discrete chirp-Fourier transform (DCFT) for parameter estimations in weak signal and high dynamic environments. Compared with the conventional receiver architecture that uses closed-loop acquisition and tracking, it is more suitable for open-loop acquisition. The proposed method combines DCFT and block zero-padding. In this way, the post-correlation signal is coherently post-integrated with the bit sequence stripped off, and the high dynamic parameters are precisely estimated using the threshold set based on a false alarm probability. In addition, the detection performance of the proposed method is analyzed. Simulation results show that compared with the BASIC method, the proposed method can precisely detect the high dynamic parameters in lower SNR when the length of the received signal is fixed.