Applied Mechanics and Materials Vols. 631-632

Abstract: This paper mainly discusses a composite model which the end time is an additive hazard function and the recurrent event process is a proportional intensity function, the covariate is time-independent, and censoring is dependent on recurrent events process and end times. Based on the likelihood method, Delta method, U-statistic method and the idea of general estimation equation, the estimation of unknown parameters and unknown functions in this composite model is proposed.

Abstract: An improved mutual information method is approached to build correlation relationships to thesauri through analysis and comparison of the problems among artificial means, co-occurrence frequency methods and mutual information method. The experimental results show that the proposed method is more objective and feasible than traditional method and it is more useful for subsequent identification.

Abstract: A single chart, instead of and R charts or and S charts, to simultaneously monitor the process mean and variability would reduce the required time and effort. A number of studies have attempted to find such charts. Moreover, a number of studies demonstrated that the adaptive control charts may detect process shifts faster than the fixed control charts. This paper proposes the EWMA loss chart with variable sample sizes and sampling intervals (VSSI) to effectively monitor the difference of process measurements and target. An example is used to illustrate the application and performance of the proposed control chart in detecting the changes in the difference of the process measurements and target. Numerical analyses demonstrated that the VSSI EWMA loss chart outperforms the fixed sampling interval EWMA average loss chart and the Shewhart joint and S charts. Therefore, the VSSI EWMA loss chart is recommended.

Abstract: In traditional Bayesian classification data mining methods, there may be defects such as predictions unreliable because the selected predictors are little or not related with the target factor. this paper analyzes the correlation between predictors and the target factor using correlation coefficient based on Bayesian classification model and combines with Hadoop distributed file system and parallel programming models to explore an improved algorithm. The experiments show that this method not only makes the prediction more reliable but also saves resources and improves the efficiency of the algorithm greatly. In addition, it is suitable for massive data processing.

Abstract: Urbanization progress and industrial structure change as an important content of economic development, the dynamic correlation effects in the process of economic development continuously. Based on Gansu province in 1978-2011 time series data, building the VAR model of urbanization progress and industrial structure change, analysis of the dynamic effect between urbanization progress and industrial structure change in Gansu province, and grasp the mutual influence of both from spatial evolution, between urbanization progress and industrial structure change is dynamic effect in Gansu province, but the impact of industrial structure change on urbanization progress is stronger than the impact of urbanization progress on industrial structure change, Suggestions are given on this basis.

Abstract: Recurrent events data refers to the observation of individuals, which contains the recurrent event time of interest. This paper mainly discusses a joint model when the end time is a multiplicable hazard function and the recurrent event process is a multiplicable intensity function. Based on the likelihood method, Delta method, U-statistic method and the idea of general estimation equation, the estimation of unknown parameters and unknown functions in the model is provided. It provides a new method of parameter estimation for the statistic analysis of recurrent events data.

Abstract: BP neural network is promising methods for the prediction of financial time series because it use a risk function consisting of the empirical error and a regularized term which is derived from the structural risk minimization principle. This study applies BP neural network to predicting the stock price index. In addition, this study examines the feasibility of applying BP neural network in financial forecasting. The experimental results show that BP neural network provides a promising alternative to stock market prediction.

Abstract: A new uncertainty information representation—hesitant three-parameter interval grey number is put forward, and its distance estimation approach of hesitant three-parameter interval grey number and sequence are studied. Based on the distance definition of hesitant three-parameter interval grey number, the paper defines the ideal optimization scheme is the positive clout and the ideal inferior scheme is the negative clout of the grey target and comprehensively considers the optimization and the positive off-target distance, and base on the spatial analysis, defines the comprehensive target distance. In the last, the model validity and practicability is demonstrated by example.

Abstract: Another novel algorithm of learning from examples is presented. A significant deference between a traditional algorithm with the new item-based algorithm is that the traditional algorithm must scan in example space (i.e. scanning the given examples one by one) to obtain the needed heuristic information, while the new item-based algorithm scans in item space (i.e. scanning items one by one) and then executes some simple calculations to obtain the same heuristic information as the traditional algorithm to do. Owing to the two facts that an item can contain thousands of examples and that the time expanded on scanning an item equals to the time on an example, the ability of the new algorithm has been revolutionarily increased, so that it can treat efficiently with the learning tasks with mass data, with which the traditional algorithms cannot deal.