Humboldt-Universität zu Berlin - High Dimensional Nonstationary Time Series

Event: Hermann Otto Hirschfeld Lecture (HOH) 2019

08.11.2019 The HOH Lecture 2019 took place on November 6 and 8, 2019. Charles F. Manski presented his work on "Patient Care under Uncertainty". The lectures draw on Manski's new book on the subject (https://press.princeton.edu/titles/30223.html) and on associated research articles.

The Lecture Series is named after the remarkable German-American statistician Hermann Otto Hartley (known as HOH). Born H.O. Hirschfeld, he completed his PhD in mathematics at Humboldt-Universität in 1934. Shortly thereafter, HOH, who was of Jewish descent, emigrated to England where he also changed his name a few years later. The Hermann Otto Hirschfeld Lectures at Humboldt-Universität are given since 2003 in memory of an excellent scientist and an advocate of quantitative methods.

Gender

16.11.2019. One of the streets in Berlin Mitte has been named after Hilda Geiringer, the famous female mathematician. Recall that IRTG 1792 regularly organises a series of lectures in her memory.

(Hilda Geiringer, 1893-1973, the the first female professor in applied mathematics in Berlin,

the second in the field of mathematics in Germany and a role model for women in science)

Read more about the renaming here.

Event: CENTRAL Workshop 2019

On November 8, 2019, the CENTRAL workshop took place at HU Berlin. It was an international networking workshop where participants from HU Berlin, Charles University, Prague and University of Warsaw, Poland came together and discussed their research ideas and future collaboration with their colleagues. The objective of the workshop was to promote joint research projects between the participating institutes. Kainat Khowaja, Vanessa Emanuela Guarino and Anna Shchekina from IRTG 1792 participated on behalf of HU Berlin. 

Journal

Professor Wolfgang Karl Härdler, speaker of the IRTG 1792 , and Professor Steven Kou of the National University of Singapore are the editors-in-chief for the new journal " Digital Finance - Smart Data Analytics, Investment Innovation, and Financial Technology ", as covered by the IRTG news last year.

This journal is a top peer-reviewed academic and practitioner journal that publishes high-quality articles with a focus on digital finance and innovation. The journal publishes theoretical or empirical, qualitative or quantitative papers of interest to academics, practitioners, and regulators with the emphasis on empirical, financial market, and investment innovation.

The Respective Website, with further information, can be reached here .

SSRN Top Ten Lists

Professor Wolfgang K. Härdle's, Speaker of the IRTG 1792, Alisa Kim's, and Simon Trimborn's paper "VCRIX - A Volatility Index for Crypto-currencies" is recently listed on: 

22.11.2019. SSRN's Top Ten download list for: PSN: Exchange Rates & Currency (Comparative) (Topic) and PSN: Exchange Rates & Currency (International) (Topic).

We featured this research outlet previously as IRTG 1792 DP 2019-027 here.

10.12.2019 The paper, "VCRIX - A VOLATILITY INDEX FOR CRYPTO-CURRENCIES", was recently listed on SSRN's Top Ten download list for: International Finance eJournal.

12.12.2019 The paper, "VCRIX - A VOLATILITY INDEX FOR CRYPTO-CURRENCIES", was recently listed on SSRN's Top Ten download list for: Econometric Modeling: Derivatives eJournal.

The Alexander Von Humboldt Research Award

This years' Friedrich Wilhelm Bessel Research Award which is one of The Alexander von Humboldt Foundation research awards has been awarded to Professor Wei Biao Wu, who visited IRTG 1792 for a short course on High Dimensional Data last year. 

The Alexander von Humboldt Foundation links Germany to the knowledge of the world. In global competition for the topmost experts, it offers to this end various programmes to attract scientists and scholars at different stages of their careers. Support involves both funding and personal supervision in all matters relating to a stay in Germany and to subsequent cooperations. For these cooperations, the Alexander von Humboldt Foundation offers numerous other kinds of sponsorship in the framework of the alumni programme. In this way, an active international network of more than 29,000 academics has grown up since the Foundation was founded in 1953.

The Humboldt Research Awards have been a central pillar of the Alexander von Humboldt Foundation’s sponsorship activities since 1972. The awards honour the academic achievements to date of internationally recognised scientists and scholars from abroad. A not insignificant number of award winners has subsequently been granted the Nobel Prize. The research award programmes take particular account of the needs of the international academic elite. The award winners are invited to spend time for research in Germany and are able to organize their stay on a flexible basis. Apart from sponsoring international science, the Alexander von Humboldt Foundation aims to develop a personal relationship between the researchers from all countries and disciplines and Germany itself. This does not only happen in laboratories and libraries, but in everyday life: through contacts to the people of the country. In order to intensify these contacts, the Alexander von Humboldt Foundation sponsors German courses for the award winners and their spouses. At the award winners’ symposium in Bamberg and the Alexander von Humboldt Foundation’s annual meeting in Berlin, there is an opportunity to get to know other Humboldtians and the Foundation’s staff personally.

AI and Data Science One Day Workshop

21.11.2019. Daniel Jacob, PhD student of the IRTG 1792, participated in a one day workshop on AI and Data Science held at the Department of Statistics, National Cheng Kung University Tainan.

The theme of the workshop revolved around the idea that AI and Data Science promise to produce solutions to many problems once provided with many data. This might not (yet) be true and one needs to understand the underlying statistics and econometric assumptions in order to evaluate such models. To shed some light into these models and put them into prospective, speakers from statistics, applied mathematics and computer science where invited to speak about deep learning, complex network analysis, fintech and machine learning in econometrics.

This one-day workshop, featured a broad range of topics where every speaker gave its insights into research and where they use AI and Data Science. About 60 listeners discussed the topics and asked questions.

We would like to thank the organizers, Ying Chen (Department of Mathematics, National University of Singapore) and Ray-Bing Chen (Department of Statistics, National Cheng Kung University) for this great workshop. It is very important that in an era of machine learning, big data and AI, everybody understands what these models and algorithms can be used for and what they can not do. We also need to rethink our understanding of “data” and on the same time always be critical about the meaning of causation.

Especially for clinical workflows, presented by Weichung Wang from National Taiwan University, such novel AI algorithms can be used to analyze images and detect e.g. early stages of prostate cancers. This leads to a higher rate of treatment and can reshape the medical examination as we know it. However, we still need and want to understand why a person shows such results and what we can do to prevent such outcomes.

Participation

04.12.2019. Wolfgang Karl Härdle, speaker of the IRTG 1792 will give a talk on 'A Dynamic Network Perspective of Cryptocurrencies' for Academic lecture series at the Department of Applied Mathematics, National Sun Yat-sen University 國立中山大學應用數學系, located in Taiwan. The abstract of the talk is given below:

Abstract

Cryptocurrencies are becoming an attractive asset class and are the focus of recent quantitative research. The joint dynamics of the cryptocurrency market yields information on network risk. Utilizing the adaptive LASSO approach, we build a dynamic network of cryptocurrencies and model the latent communities with a dynamic stochastic block model.We develop a dynamic covariate-assisted spectral clustering method to uniformly estimate the latent group membership of cryptocurrencies consistently. We show that return inter-predictability and crypto characteristics, including hashing algorithms and proof types, jointly determine the crypto market segmentation. Based on this classification result, it is natural to employ eigenvector centrality to identify a cryptocurrency’s idiosyncratic risk. An asset pricing analysis finds that a cross-sectional portfolio with a higher centrality earns a higher risk premium. Further tests confirm that centrality serves as a risk factor well and delivers valuable information content on cryptocurrency markets.

Participation

04.12.2019. Cathy Yi-Hsuan Chen, participating faculty at IRTG 1792, will give a talk on 'SONIC: Social Network with Influencers and Communities' for Academic lecture series at the Department of Applied Mathematics, National Sun Yat-sen University 國立中山大學應用數學系, located in Taiwan. The abstract of her talk is given below:

Abstract

The integration of social media characteristics into an econometric framework requires modeling a high dimensional dynamic network with dimensions of parameter Θ typically much larger than the number of observations . To cope with this problem, we introduce a new structural mode SONIC which assumes that (1) a few influencers drive the network dynamics; (2) the community structure of the network is characterized as the homogeneity of response to the specific infuencer, implying their underlying similarity. An estimation procedure is proposed based on a greedy algorithm and LASSO regularization. Through theoretical study and simulations, we show that the matrix parameter can be estimated even when the observed time interval is smaller than the size of the network . Using a novel dataset retrieved from a leading social media platform StockTwits and quantifying their opinions via StockTwits and quantifying their opinions via natural natural language processing, we model the opinions network language processing, we model the opinions network dynamics among a select group of users and further among a select group of users and further detect the latent communities. With a sparsity With a sparsity regularization, we can identify important nodes in the regularization, we can identify important nodes in the network.

Participation

05.12.2019. Wolfgang Karl Härdle, speaker of the IRTG 1792 gave a talk on 'Influencers and Communities in Social Networks' at the Institute of Statistical Science, Academia Sinica (中央研究院統計科學所究所), located in Taiwan. The abstract of the talk is given below:

Abstract

The integration of social media characteristics into an econometric framework requires modeling a high dimensional dynamic network with dimensions of parameter Θ typically much larger than the number of observations . To cope with this problem, we introduce a new structural mode SONIC which assumes that (1) a few influencers drive the network dynamics; (2) the community structure of the network is characterized as the homogeneity of response to the specific infuencer, implying their underlying similarity. An estimation procedure is proposed based on a greedy algorithm and LASSO regularization. Through theoretical study and simulations, we show that the matrix parameter can be estimated even when the observed time interval is smaller than the size of the network . Using a novel dataset retrieved from a leading social media platform StockTwits and quantifying their opinions via StockTwits and quantifying their opinions via natural natural language processing, we model the opinions network language processing, we model the opinions network dynamics among a select group of users and further among a select group of users and further detect the latent communities. With a sparsity With a sparsity regularization, we can identify important nodes in the regularization, we can identify important nodes in the network.

Participation: First Yushan Conference

06.12.2019. The First Yushan Conference took place on December 6, 2019 at NCTU in Hsinchu, Taiwan. Wolfgang K. Härdle, Speaker of the IRTG 1792, gave a talk on “CRIX - a CRypto IndeX”.

The IRTG PhD students Elena Ivanova, Georg Keilbar and Bingling Wang also presented their work at the conference.

Highlight

18.12.2019. The funding proposal for the Blockchain Research Center (BRC) of Prof. Wolfgang Karl Härdle, speaker of IRTG 1792, was appproved under the First Joint Seed Funding call between the Humboldt-Universität zu Berlin and the Universität Zürich. 

Joint Seed Money Funding

As part of their strategic partnership, the University of Zurich and Humboldt-Universität zu Berlin issued their first joint seed funding call to provide financing for collaborations between the two universities in the fall semester 2019.