Identification and Categorization of Climate Change Risks

The scientific evidence that climate is changing due to greenhouse gas emission is now incontestable, which may put many social, biological, and geophysical systems in the world at risk. In this paper, we first identified main risks induced from or aggravated by climate change. Then we categorized them applying a new risk categorization system brought forward by Renn in a framework of International Risk Governance Council. We proposed that ＂uncertainty＂ could be treated as the classification criteria. Based on this, we established a quantitative method with fuzzy set theory, in which ＂confidence＂ and ＂likelihood＂, the main quantitative terms for expressing uncertainties in IPCC, were used as the feature parameters to construct the fuzzy membership functions of four risk types. According to the maximum principle, most climate change risks identified were classified into the appropriate risk types. In the mean time, given that not all the quantitative terms are available, a qualitative approach was also adopted as a complementary classification method. Finally, we get the preliminary results of climate change risk categorization, which might iay the foundation for the future integrated risk management of climate change.

Research on Power Consumption Anomaly Detection Based on Fuzzy Clustering and Trend Judgment

Among the end-users of the power grid,especially in the rural power grid,there are a large number of users and the situation is complex.In this complex situation,there are more leakage caused by insulation damage and a small number of users stealing electricity.Maintenance staff will take a long time to determine the location of the abnormal user meter box.In view of this situation,themethod of subjective fuzzy clustering and quartile difference is adopted to determine the partition threshold.The power consumption data of end-users are divided into three regions:high,normal and low,which can be used to screen users in the area of abnormal power consumption.Then the trend judgment method is used to further accurately screen to improve the accuracy and reduce the number of users in the abnormal range.Finally according to abnormal power consumption auxiliary locate abnormal electricity users list box.Then the simulation environment is set to verify the application of membership fuzzy clustering and trend judgment in power consumption data partition.

Exponential Fuzzy C-Means for Collaborative Filtering

Collaborative filtering （CF） is one of the most popular techniques behind the success of recommendation system. It predicts the interest of users by collecting information from past users who have the same opinions. The most popular approaches used in CF research area are Matrix factorization methods such as SVD. However, many well- known recommendation systems do not use this method but still stick with Neighborhood models because of simplicity and explainability. There are some concerns that limit neighborhood models to achieve higher prediction accuracy. To address these concerns, we propose a new exponential fuzzy clustering （XFCM） algorithm by reformulating the clustering＇s objective function with an exponential equation in order to improve the method for membership assignment. The proposed method assigns data to the clusters by aggressively excluding irrelevant data, which is better than other fuzzy C-means （FCM） variants. The experiments show that XFCM-based CF improved 6.9% over item-based method and 3.0% over SVD in terms of mean absolute error for 100 K and 1 M MovieLens dataset.