Obtaining crowd flow distribution with recognized human intention is extremely valuable for a series of applications for metropolitan cities. Previous solutions look at spatial correlation and temporal periodicity based on historical crowd flow information to calculate future crowd flow distribution. However, these mechanisms cannot recognize the intention behind crowd flow. We address this problem by leveraging a key insight – people's intention behind their movement is highly correlated with the point-of-interest (POI) distribution of the corresponding regions and adjacent regions. Therefore, we proposeDeepFlowGento model the complicated relationship between crowd flow, POI, check-ins, and time to generate intention-aware crowd flow. Specifically, we solve the conflict between dynamic crowd flow and static POI distribution by fusing the information in both time and POI domains. Besides, we employ a sequence of residual blocks inDeepFlowGento address the challenges of modeling the diverse temporal rhythms and heterogeneous influence of POI. Furthermore, we examine the generated intention-aware crowd flow from two aspects to substantiate the reasonability ofDeepFlowGen. Extensive experiments demonstrate that our model outperforms the state-of-the-art solutions by at most 30 percent in terms of NRMSE of total crowd flow. Moreover, the correlation between the generated intention-aware crowd flow and the check-in distribution across different categories of POIs is as high as 0.90 and 0.80 in Beijing and Shanghai. Combined with extensive case studies, we demonstrate the strong ability of our model in generating intention-aware crowd flow.

Flexibility and evolution have been a hot topic in the context of business process management. However, contemporary process mining techniques assume processes to be in a steady state, which will result in a mixed-model mined from the event log. Business process drift detection is a family of methods to detect changes by analyzing the event log, but existing methods have some disadvantages in dealing with concept drifts. First, most of these methods detect changes depending on an exploration of a potentially large feature space and are time consuming. Second, with the size of delay period becoming small, the accuracies of some methods will drop rapidly. In this article, we address these problems by proposing a novel drift detection technique called LCDD that significantly differs from existing methods. The core idea is to find out new features and disappeared features from traces after the log fragment meets local completeness. To begin with, we use the relations of direct succession as the lightweight feature, which are compared between two windows (complete window and detection window). Then, we find new direct successions and stable disappeared direct successions by just moving detection window. Finally, the forgetting mechanism is used to abandon some direct successions after finding a change point. An extensive empirical evaluation shows that LCDD is faster and more accurate.

Event logs are the basis of process mining techniques and tools that extract process behavior information for better understanding and optimization of business processes. While it has been widely realized that the degree of completeness of event logs may largely determine the effectiveness of these techniques, how to estimate the completeness of event logs has not yet been fully addressed. This is mainly because ground-truth process models are usually unknown. To attack this problem, we pay a closer look to several concepts and implicit assumptions in the log completeness estimation problem and characterize it as a special case of the species estimation problem in the field of statistics. Although species estimation is still an open problem, a number of statistic models and techniques with approximate solutions have been available. To investigate the relevance of these methods for event log completeness estimation, we have designed and conducted a wide scope of empirical study and quantitative experiments on both real-world and synthesized event logs to compare the performance of these methods. In addition, the completeness estimation of several important and widely used real-world events logs are reported for the first time together with some best practice experience learned through this research.

Process mining links traditional model-driven Business Process Management and data mining by means of deriving knowledge from event logs to improve operational business processes. As an impact factor of the quality of process mining results, the degree of completeness of the given event log should be necessarily measured. In this paper an approach is proposed in the context of mining control-flow dependencies to evaluate the local completeness of an event log without knowing any information about the original process model. Experiment results show that the proposed approach works robustly and gives better estimation than approaches available.

A multilayered network with the first layer consisting of parabolic neurons (constrained second-order neurons) is proposed. Each parabolic neuron requires onlyN+2 interconnections (where N is the number of inputs), which is virtually the same as a linear neuron. Such a network needs fewer neurons or layers than standard backpropagation (BP) to solve the same problem, and it converges much faster. The training is achieved through the backpropagation learning law. It is concluded that incorporating parabolic neurons onto the BP network can significantly improve the network's performance without resulting in an explosive number of interconnection weights

A modified learning algorithm for BP (backpropagation) neural networks is presented based on the interpretation of a neuron in an (N+1)-dimensional space,RN+1, and the analysis of how a multilayered network performs a classification task by collective use of the B (boundary) neurons in the first layer. The role of B neurons and L (logic) neurons is discussed. A B neuron represents a linear boundary in the input space RN. An L neuron in the second layer defines in RNa convex piecewise linear boundary that is formed by a set of line segments corresponding to connected B neurons. An L neuron in the third layer defines in RNa complicated boundary that can have both convex parts and concave parts. Each subboundary of it corresponds to an L neuron in the second layer. The nonlinear function does not change the structure of a boundary but smooths the angular vertices of a boundary. It is shown that for the two-class problems with convex boundaries, the nonlinear function in an L neuron can be replaced with a step and the weights of the L neuron can be fixed. Computer simulation shows that this algorithm can learn more quickly than the ordinary BP algorithm, even in cases where the ordinary BP algorithm will fail