【6月25日】统数学院2016手拉手Workshop系列之（三）：Spatial-temporal time series and image analysis
发布日期：2016-06-21主题：Spatial-temporal time series and image analysis
时间：2016年6月25（星期六）9：00-17：00
地点：沙河校区，4号楼107
具体日程安排如下：
09：00-10：30 Professor Chunming Zhang, University of Wisconsin
Title: New Statistical Methods for Analyzing Complex Imaging Data
The course introduces new statistical learning and inference methods in the analysis of large-scale complex imaging data. This lecture intends to discuss issues, algorithms, challenges, recent developments, and open problems.
Topics to be covered:
1. Semiparametric estimation and inference for spatial-temporal brain fMRI data.
2. Large-scale simultaneous inference and controlling FDR for spatial imaging data.
3. Estimating large error covariance matrices for temporally correlated data.
4. Statistical methods for neuronal functional connectivity.
10:40 - 12:00 Dr. Yanfei Kang, BaiduInc
Title: Exploring time series collections used for forecasting evaluation
It is common practice to evaluate the strength of forecasting methods using collections of well-studied time series datasets, such as the M3 data. But how diverse are these time series, how challenging, and do they enable us to study the unique strengths and weaknesses of different forecasting methods? In this paper we propose a visualisation method for a collection of time series that enables a time series to be represented as a point in a 2-dimensional instance space. The effectiveness of different forecasting methods can be visualised easily across this space, and the diversity of the time series in an existing collection can be assessed. Noting that the M3 dataset is not as diverse as we would ideally like, this paper also proposes a method for generating new time series with controllable characteristics to fill in and spread out the instance space, making generalisations of forecasting method performance as robust as possible.
14:00 - 15:30 Professor Zhengjun Zhang, University of Wisconsin
Title: Practical extreme value analysis methodology and case studies
Extreme value theory is concerned with describing the extreme values of an observed process and the predictions of future extremes in the process. In practice, you cannot rely on statistical modeling by normal, lognormal, Weibull, or many other commonly used distributions all the way out into extreme tails. This lecture will give an overview of practical extreme value analysis methodology and case studies in precipitation data.
References:
1, An Introduction to Statistical Modeling of Extreme Values, by Stuart Coles, Springer, 2001.
2, Extremes in Nature, An Approach Using Copulas, by GianfaustoSalvadori, Carlo De Michele, Nathabandu T. Kottegoda, Renzo Rosso, Springer, 2007.
3, Statistics of Extremes, with Application in Environment, Finance and Insurance, by Richard Smith, in Extreme Values in Finance, Telecommunications, and the Environment edited by BarbelFinkenstadt, HolgerRootzen, CRC Press, 2003.
4, Max-autoregressive and Moving Maxima Models for Modeling Extremes, by Zhengjun Zhang, Liang Peng, and Timothy Idowu, in Extreme Value Modeling and Risk Analysis: Methods and Applications, Editors: DipakDey and Jun Yan. Chapman Hall/CRC, 2015.
5, Dependence Modeling with Copulas, By Harry Joe, Chapman Hall/CRC, 2014.
15:40 - 17:00 Professor Lu Deng, Central University of Finance and Economics
Title: Tail Dependent Pattern Recognitions in China's Smog Extreme Co-movements
[编辑]：孙颖
