Search

With the continuous improvement of international "pharmacovigilance" technology and methods,it becomes the key part of the post-marketing evaluation. This issue is based on this research background, and also means to find out the Chinese medicine safety monitor which consistents with the reality. A common problem is that those who choose a career in pharmacovigilance know how the complex data presented to us are a source of both fascination and frustration. In the 70's, for the first time data mining technology in the international pharmacovigilance turn up, we try to establish new signal detection method to make contributes to post-marketing evaluation of Chinese medicine and establishment of registration. Building the national adverse reaction reporting database is widely used in western country. Nature of the problem is that pharmacovigilance issues can come through a lot of assumptions into the statistical problems, different assumptions are for different statistical tests. Through the traditional imbalance between the proportion of fourfold table for other assumptions, few countries use in practice, this does not involve evidence, but this issue provides the introduce of the principle. Methods include the ratio of the report of the Netherlands (ROR), the proportion of reports than the UK ratio (PRR),WHO's information points (IC), the U.S. Food and Drug Administration empirical Bayes (EBS), etc. Because there is no international gold standard of the signal detection method, at first we use the simulation comparing these four methods of data mining, From the point of specificity, the sample size demand, this issue views the advantages and disadvantages of four methods and application conditions,and from a technical point of view and try to propose a new signal detection method, for example, Hierarchical Bayesian.

Clustering algorithms are important methods widely used in mining data streams because of their abilities to deal with infinite data flows. Although these algorithms perform well to mining latent relationship in data streams, most of them suffer from loss of cluster purity and become unstable when the inputting data streams have too many noisy variables. In this article, we propose a clustering algorithm to cluster data streams with noisy variables. The result from simulation shows that our proposal method is better than previous studies by adding a process of variable selection as a component in clustering algorithms. The results of two experiments indicate that clustering data streams with the process of variable selection are more stable and have better purity than those without such process. Another experiment testing KDD-CUP99 dataset also shows that our algorithm can generate more stable result. [ABSTRACT FROM AUTHOR]

In this article, we introduce the L1 regularized support vector machine (L1-SVM) as an effective feature selection technique into the modeling of financial early warning system (EWS), for the purpose of establishing compact financial EWSs for Chinese-listed companies. By introducing LASSO penalty into the SVM framework, the L1-SVM is a capable methodology to select causative features in classification problem. We evaluate the feature selection performance of L1-SVM under different circumstances through numerical simulations and find it suitable for selecting features for financial EWS. In the real study, we establish four financial EWSs with features selected by L1-SVM and compare them with those trained with full features. The empirical result illustrates that our EWSs, with only minority of features, outperform significantly than full ones in the respect of generalization performance, which indicates the feasibility of L1-SVM in real applications. Copyright © 2014 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]