New Discovery of the Late Triassic Terrigenous Sediments in the Great Xing＇＇an Range Region,NE China and its Geological Significance

Objective The Great Xing＇an Range is located in the eastern section of Central Asian Orogenic Belt（CAOB）.As a superposed position of multiple tectonic domains,its structural evoIlution has always been a focused issue of geological research.

Study on the Determination of the Significant National Seismic Monitoring and Protection Regions

The paper describes firstly the principles and scientific train of thought involved in determining the significant seismic monitoring and protection regions (SSMPR) in China. The principles include the gradation principle, i.e. the national level SSMPR and the provincial level SSMPR, the principle of highlighting priorities, namely, the area of an SSMPR should be a fraction of the total area of the country or of the respective province, but the earthquake losses incurred in SSMPR should be a major proportion of the national or provincial ones. The scientific train of thought adopted is to determine the SSMPR on the basis of seismic hazard assessment and loss estimation. Secondly, it reviews the achievements in determining the SSMPRs for the period from 1996 to 2005. The result shows that 10 strong earthquakes occurred during that period in the areas with earthquake monitoring and prediction capability available on the Chinese continent, 8 of which occurred in SSMPRs with the economic loss and death toll accounting for 67% and 92% of the total loss on the Chinese mainland. Lastly, the paper introduces preparatory research for determining the SSMPR for the period from 2006 to 2020, including decade-scale mid-and long-range seismic risk assessment based on seismology, seismogeology, geodesy, earthquake engineering, sociology and stochastics and so on, and the national seismic risk probability map, the seismic hazard (intensity) map, earthquake disaster losses map and the comprehensive seismic risk index, etc. obtained for the period of 2006 to 2020.

Conclusions on the Implementation of Regulation of the National Significant Seismic Monitoring and Protection Regions from 1996 to 2012 in the Chinese Mainland

The regulation of the National Significant Seismic Monitoring and Protection Regions(NSSMPR for short) is defined by the Law of the Peoples Republic of China on Protecting Against and Mitigating Earthquake Disasters.The first stage of implementation of the regulation of NSSMPR in the Chinese mainland was finished from 1996 to 2005.The second stage is being carried on from 2006 to 2020.With the support of the National Social Science Foundation,this paper follows up and evaluates the implementation of the regulation of NSSMPR from 1996 to 2012 in the Chinese mainland.Based on analysis of earthquake examples and investigation data,we find that the effect of disaster mitigation is good,and on this basis,some suggestions are proposed to improve the regulation of NSSMPR.

NLWSNet:a weakly supervised network for visual sentiment analysis in mislabeled web images

Large-scale datasets are driving the rapid developments of deep convolutional neural networks for visual sentiment analysis.However,the annotation of large-scale datasets is expensive and time consuming.Instead,it iseasy to obtain weakly labeled web images from the Internet.However,noisy labels st.ill lead to seriously degraded performance when we use images directly from the web for training networks.To address this drawback,we propose an end-to-end weakly supervised learning network,which is robust to mislabeled web images.Specifically,the proposed attention module automatically eliminates the distraction of those samples with incorrect labels bv reducing their attention scores in the training process.On the other hand,the special-class activation map module is designed to stimulate the network by focusing on the significant regions from the samples with correct labels in a weakly supervised learning approach.Besides the process of feature learning,applying regularization to the classifier is considered to minimize the distance of those samples within the same class and maximize the distance between different class centroids.Quantitative and qualitative evaluations on well-and mislabeled web image datasets demonstrate that the proposed algorithm outperforms the related methods.