No abstract

Mining influence in evolving entities is an important but challenging task, partly due to complex nature of it. In this paper, we mainly focus on the following problems on it with respect to stock market: (1) How to identify pairs of stocks that influence one another; (2) How to quantify the influence and capture group effects and dynamic nature of influence of each stock; (3) How to adopt approximate approaches so that we can improve the efficiency of the proposed model. To tackle these problems, a novel graph-based mining method, which utilizes time series and volume information collaboratively is proposed, and several optimized algorithms are presented. Besides, two extended metrics to capture the dynamic and group nature of influence based on the model are derived. Furthermore, we also suggest a potential application of the model to stock price prediction. The experimental results on both synthetic and real data sets verify the effectiveness and efficiency of our approach. Some insights on this paper can be the ideas of analyzing the influence of evolving entities using the social network analysis methods.

No abstract

No abstract

This paper aims to provide a thorough study on the effectiveness of the transformation-based ensemble defence for image classification and its reasons. It has been empirically shown that they can enhance the robustness against evasion attacks, while there is little analysis on the reasons. In particular, it is not clear whether the robustness improvement is a result of transformation or ensemble. In this paper, we design two adaptive attacks to better evaluate the transformation-based ensemble defence. We conduct experiments to show that 1) the transferability of adversarial examples exists among the models trained on data records after different reversible transformations; 2) the robustness gained through transformationbased ensemble is limited; 3) this limited robustness is mainly from the irreversible transformations rather than the ensemble of a number of models; and 4) blindly increasing the number of sub-models in a transformation-based ensemble does not bring extra robustness gain.