Joint QF-STAT PhD Webinars
Zoom Link and Password: Request by e-mailing to Mrs. Laurie Dalessio (firstname.lastname@example.org)
Spring 2022 Semester
Liting Chiang | 11 February 2022 | TBD
Speaker: Liting Chiang
Date/Time: 11 Feb. 2022, Friday, 10:00 AM - 11:00 AM
Rui Ding | 25 February 2022 | TBD
Speaker: Rui Ding
Date/Time: 25 Feb. 2022, Friday, 10:00 AM - 11:00 AM
Eric Werneburg | 11 March 2022 | Training Neural Networks Using Reproducing Kernel Space Interpolation
Speaker: Eric Werneburg
Title: Training Neural Networks Using Reproducing Kernel Space Interpolation
Date/Time: 11 Mar. 2022, Friday, 10:00 AM - 11:00 AM
Abstract: We introduce and study the theory of training neural networks using interpolation techniques from reproducing kernel Hilbert space theory. We generalize the method to Krein spaces, and show that widely-used neural network architectures are subsets of reproducing kernel Krein spaces. We introduce the concept of “associated Hilbert spaces” of such conventional neural networks. Using such machinery results in networks with the same expressivity, but which tend to generalize better after being trained. Next, we introduce Agler’s theorem in the context of neural network theory to show that there is one activation function, the expressivity of which is superior to all other desirable activation functions. Finally, we show the successful application of our theory in the field of neurourology and urodynamics.
TBD | 25 March 2022 | TBD
Date/Time: 25 Mar. 2022, Friday, 10:00 AM - 11:00 AM
TBD | 8 April 2022 | TBD
Date/Time: 8 Apr. 2022, Friday, 10:00 AM - 11:00 AM
El Mehdi Ainasse | 22 April 2022 | TBD
Speaker: El Mehdi Ainasse
Date/Time: 22 Apr. 2022, Friday, 10:00 AM - 11:00 AM
Fall 2021 Semester
Teng Chen | 10 September 2021 | Tail Index Estimation Based on CVaR
Speaker: Teng Chen
Title: Estimation of Heavy Tails
Date/Time: 10 Sep. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: In extreme value theory, the tail index is a critical measure for assessing the heavy-tailed behavior of a distribution. In this talk, we propose a new class of Pickands-type estimators based on conditional value-at-risk (CVaR) for the tail index. The CVaR-based estimators not only inherit the universal consistency, the location and scale invariant properties of Pickands estimator but also significantly improve the asymptotic efficiency and reduce the sensitivity to the choice of the intermediate order statistics. We prove the weak and strong consistency, and the asymptotic normality of these estimators under quite general conditions. A detailed investigation of the asymptotic behavior of the estimators in the family is carried out in order to identify which members of the family perform optimally. A simulation study involving a wide range of distributions shows the CVaR-based estimators have good finite-sample performance and compare favorably with the other estimators.
Jiaju Miao | 24 September 2021 | Sector Rotation Strategy via Machine Learning
Speaker: Jiaju Miao
Title: Sector Rotation Strategy via Machine Learning
Date/Time: 24 Sep. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: Asset-specific factors have been widely used to explain financial returns and measure asset-specific risk premia. In this paper, we employ these factors in various machine learning models to measure sector risk premia. We initially start with a comparison of the prediction of different models and demonstrate large economic gains from using machine learning for sector forecasting. We develop an assembling algorithm that combines different models based on the history of their performance, and we prove that the risk premia of the resulting meta-algorithm are not much worse than the risk premia of the infeasible optimal assembling. The resulting meta-strategy is used for portfolio investment that substantially outperforms the market.
Qinyun Zhao | 8 October 2021 | Learning and control of time-varying linear quadratic regulator
Speaker: Qinyun Zhao
Title: Learning and control of time-varying linear quadratic regulator
Date/Time: 8 Oct. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: Economic agents are hard to obtain complete information about all the parameters which could affect the payoffs that they receive. However, they could get feedback from observing the rewards of their actions. There is a voluminous literature on learning and control with unknown parameters. In this talk, we first introduce the work in Wieland's paper about his numeric method to get optimal control in a linear process with unknown parameters. After that, I would like to show my on-going work of applying a more efficient method, rollout algorithm to obtain approximate optimal control under more complicate time-varying environments.
Xiaoqi Dong | 22 October 2021 | Survey on multi-agent reinforcement learning
Speaker: Xiaoqi Dong
Title: Survey on multi-agent reinforcement learning
Date/Time: 22 Oct. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: In recent years, reinforcement learning (RL) has been witnessed to have magnificent advances. Most of the RL real-world applications, however, involve the participation of more than one agent, and fall into the realm of multi-agent RL (MARL). Although multi-agent learning (MAL) was first discussed in 1999 and is relatively less popular in 2000s, due to advances in single-agent RL (SARL) techniques and improvements of computation performance, this realm has recently re-merged. In this survey, we pro- vide a selective overview of algorithms in MARL. In specific, we review the algorithms in two representative frameworks, Markov games and extensive-form games, in accordance with the types of learning goals, which are fully cooperative, fully competitive, and a mix of the two. (This talk is a review of Zhang, Kaiqing, Zhuoran Yang, and Tamer Başar. "Multi-agent reinforcement learning: A selective overview of theories and algorithms." Handbook of Reinforcement Learning and Control(2021): 321-384).
Ryan Kaufman and Prof. James Glimm | 5 November 2021 | GA1.2: An Intraday and Overnight Risk Model and Interest Rate Risk
Speaker: Ryan Kaufman and Prof. James Glimm
Title: GA1.2: An Intraday and Overnight Risk Model and Interest Rate Risk
Date/Time: 5 Nov. 2021, Friday, TBD
Abstract: Both daily and intraday risk models miss one of the key elements of a day's risk: the overnight jump. Intraday risk models will underestimate the day's risk when using time series analysis without a model for the overnight step, during which new information becomes available to traders, but prices freeze, a feature not captured in the time series representation of the data. Daily risk models will also underestimate the risk. Daily time series does capture the overnight step within the data, but averages it in with the risk of the whole day. Interest rate forecasting also comes with some caveats for modeling, most notably is the effect of political actions. We incorporate this into our model by including information about the present and future federal reserve rate into the model for bond prices, which is used to calculate risk. We show models for both of these phenomena and present some results.
Greeshma Balabhadra | 19 November 2021 | Breakpoints in High Frequency Volatility
Speaker: Greeshma Balabhadra
Title: Breakpoints in High Frequency Volatility
Date/Time: 19 Nov. 2021, Friday, 10:00 AM - 11:00 PM
Abstract: We introduce two new high-frequency volatility estimators that account for possible breakpoints in the spot volatility process. They are l1-penalized versions of classical estimators—the quadratic variation, and, jump robust version of it, bipower variation. We show that in the presence of a mean-square error of order op(1) achieved by these classical estimators, detecting breakpoints using the volatility estimator is asymptotical equivalent to detecting them using the infeasible (latent) volatility path. The proposed estimators are evaluated in simulations and on real data. They are fast in computations, and they accurately detect breakpoints that are close to the end of the sample—both properties are very desirable for financial applications. In terms of out-of-sample volatility prediction, qualitatively, the new estimators provide more smooth and realistic forecasts; quantitatively, they outperform the aforementioned competitors and their extensions at various frequencies and forecasting horizons.
Jin Huang | 3 December 2021 | Convergence analysis of Matural policy gradient algorithms
Speaker: Jin Huang
Title: Convergence Analysis of Natural Policy Gradient Algorithms
Date/Time: 3 Dec. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: When solving the problem of reinforcement learning, we often use optimization methods such as gradient descent. In this study, we are committed to studying the convergence of the Natural Policy Gradient under multiple iterations. Besides, we continue to discuss the possibility of convergence based on Natural Actor-Critic algorithm. The purpose of the research is to achieve faster convergence and theoretically establish the premise of high algorithm efficiency.
Spring 2021 Semester
Mr. Gabriel Mihalache | 12 February 2021 | Sustainable Debt with Long Run Risk
Speaker: Gabriel Michalache
Title: Sustainable Debt with Long Run Risk
Date/Time: 12 Feb. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: We propose a parsimonious framework for the quantitative evaluation of sustainable debt in emerging markets, by embedding rich fiscal rules and a pricing kernel consistent with key stylized facts, as in Bansal and Yaron (JF, 2004), in the slow-moving debt crisis model of Lorenzoni and Werning (AER, 2019). The model can replicate the importance of a common global factor in spreads and the slow response of primary deficits to debt overhang. The framework supports the design of robust fiscal rules.
Speaker Bio: Gabriel Mihalache is an Assistant Professor of Economics at Stony Brook University since 2016. He was awarded a MSc in Applied Economics in Statistics by Clemson University in 2010 and a PhD in Economics in 2016 by the University of Rochester. His work addresses the causes and consequences of sovereign default risk for the maturity structure of sovereign debt, capital accumulation, and monetary policy.
Mr. Kevin Maritato | 26 February 2021 | An Introduction to Buffered Probability Distribution Functions
Speaker: Kevin Maritato
Title: An Introduction to Buffered Probability Distribution Functions
Date/Time: 26 Feb. 2021, Friday, 10:00AM - 11:00AM
Abstract: In this paper we introduce the concept of Buffered Probability Distribution Functions (bPDFs), as the derivative of the inverse of the Conditional Value at Risk (CVaR). We first explore the connection between this value and risk measures such as the Buffered Probability of Exceedance (bPoE). Then we describe three means of calculating the bPDF for a distribution, depending on what is known about the distribution. For functions with a closed-form formula for bPoE, we derive closed-form formulae for bPDF. We also examine using these formulae for parameter estimation using the maximum likelihood method. We then describe a method for deriving a formula for bPDF based on a numerical bPoE (calculated using Portfolio Safeguard software) when there is a closed-form formula for CVaR but no formula for bPoE. Finally, we give a method for numerically calculating bPDF for an empirical distribution, and compare the results of this method to those given by the other methods for known distributions.
Haoran Jiang| 26 March 2021 | Modeling and computation of multi-step batch testing for infectious diseases
Speaker: Haoran Jiang
Title: Modeling and computation of multi-step batch testing for infectious diseases
Date/Time: 26 Mar. 2021, Friday, 10:00AM - 11:00 AM
Abstract: We propose a mathematical model based on probability theory to optimize COVID-19 testing by a multi-step batch testing approach with variable batch sizes. This model and simulation tool dramatically increase the efficiency and efficacy of the tests in a large population at a low cost, particularly when the infection rate is low. The proposed method combines statistical modeling with numerical methods to solve nonlinear equations and obtain optimal batch sizes at each step of tests, with the flexibility to incorporate geographic and demographic information. In theory, this method substantially improves the false positive rate and positive predictive value as well. We also conducted a Monte Carlo simulation to verify this theory. Our simulation results show that our method significantly reduces the false negative rate. More accurate assessment can be made if the dilution effect or other practical factors are taken into consideration. The proposed method will be particularly useful for the early detection of infectious diseases and prevention of future pandemics. The proposed work will have broader impacts on medical testing for contagious diseases in general.
Oluyemi Oyeniran | 9 April 2021 | The Non-Clinical Statistics
Speaker: Oluyemi Oyeniran
Panelists: Fanni Zhang, Jyh-Ming Shoung, Thomas Bradstreet, Wei Zhao (Detail information)
Title: The Non-Clinical Statistics
Date/Time: 9 Apr. 2021, Friday, 10:00 AM - 11:00 AM
Abstract: We will use this opportunity to introduce to the students the drug development process and how non-clinical statistics are involved in each and every step. After the presentation, we hope the students will have a better understanding of different career paths in the pharmaceutics industry.
Mr. Mitch Gao | 23 April 2021 | NIG Process and Its Application in Option Pricing and Interest Rate Models
Speaker: Mr. Mitch Gao
Title: NIG Process and Its Application in Option Pricing and Interest Rate Models
Date/Time: 26 Mar. 2021, Friday, TBD
Abstract: NIG type time-inhomogeneous Levy process, introduced by Barndorﬀ-Nielsen as a Normal variance-mean mixture with an Inverse Gaussian mixing distribution, have greater ﬂexibility and can model a variety of distributions, making this process one of great interest from a mathematical ﬁnance point of view. In this work, we study the NIG option pricing under Esscher transformation and explore its application in term structure of interest rate models.
Yizhou Li | 7 May 2021 | Dynamic management of loan portfolios
Speaker: Yizhou Li
Title: Dynamic management of loan portfolios
Date/Time: 7 May 2021, Friday, 10:00 AM - 11:00 AM
Abstract: Dynamic management of loan portfolio is a very important task of bank. A bank can run its capital to investment loans dynamically by stochastic optimization model based on its constraint of asset and liability. At the first step of this task, we improved the accuracy of loan default prediction from LendingClub, a peer-to-pear lending platform, and this default prediction is considered as a binary classification problem. We propose a two-step training model. This model first utilizes feature-wise spline logistic regression with polynomial function to transform features, and then predicts default by logistic regression. We used metrics including the area under the ROC curve (AUC) and accuracy rate for evaluation of this algorithms, and this has been done both in-sample and out-of-sample. The feature-wise spline regression can also be used for feature selection and demonstrate visually the relationship between probability of default and feature. As a result, rather than using massive features with neural network, random forest or any other machine learning classifiers, a simple logistic regression model with a limited number of features will output the best metrics.
Fall 2020 Semester
Mr. Teng Chen | 11 September 2020 | Classification and Severity Progression Measure of COVID-19 Patients Using Proteomic and Metabolomic Sera
Speaker: Mr. Teng Chen (joint presentation with Prof. Pawel Polak and Prof. Stan Uryasev)
Title: Classification and Severity Progression Measure of COVID-19 Patients Using Proteomic and Metabolomic Sera
Date/Time: 11 Sep. 2020, Friday, 10:00 - 11:00 AM EST
Abstract: Early detection and effective treatment of severe COVID-19 patients remain one of the major challenges during the current pandemic. Analysis of molecular changes in blood samples of severe patients is one of the promising approaches to this problem. From 75 most relevant proteomic and metabolomic factors selected by Shen et al. (2020), we identify several pairs of factors (some of them after additional nonlinear spline transformation) that are highly effective in classifying severe COVID-19 cases. The performance of these pairs is evaluated in-sample, in a cross-validation exercise, and in an out-of-sample analysis on an independent dataset. Our findings can help medical experts to identify small groups of biomarkers that can be used to construct a cost-effective, short-term test for patients screening and a measure of severity progression.
Mr. Pengzhan Guo | 9 October 2020 | Customizable Career Path Recommendation with Multi-Criteria Stochastic Optimization
Speaker: Mr. Pengzhan Guo
Title: Customizable Career Path Recommendation with Multi-Criteria Stochastic Optimization
Date/Time: 9 October 2020, Friday, 10:00 - 11:00 AM EST
Abstract: Career mobility forecasting and recommendation are important research topics in talent management and people analytics. While existing models mainly focus on short-term one-period recommendation and prediction for job seekers, the long-term customizable career path recommendation (CPR) is a topic that has not been sufficiently investigated. In this project, we propose a generalized model of CPR and investigate related mathematical and economic characteristics from a long-term customizable perspective. The new objective function is capable of delivering personalized long-term career plans while upholding properties corresponding to three common facts among existing literature on job mobility. We then develop an efficient and effective searching algorithm to achieve the optimal solution of CPR based on simulated annealing and Markov chain techniques. We not only show the effectiveness of our method theoretically but also compare it to three different versions of the greedy method that includes the current state-of-the-art method empirically in addressing similar problems. From simulations based on a large real-world dataset, we show consistent results reaffirming the superior performance of our method in recommending optimal career paths based on user-defined criteria.
Mr. Yuanchen Huang | 23 October 2020 | Time Varying Parameter Selection with Spike and Slab Prior
Speaker: Mr. Yuanchen Huang
Title: Time Varying Parameter Selection with Spike and Slab Prior
Date/Time: 23 October 2020, Friday, 10:00 - 11:00 AM EST
Mr. Ziheng Chen | 6 November 2020 | On Reinforcement Learning and Compressed Sensing
Speaker: Mr. Ziheng Chen
Title: On Reinforcement Learning and Compressed Sensing
Date/Time: 6 November 2020, Friday, 10:00 - 11:00 AM EST
Xiaoqi Dong | 20 November 2020 | Weighted reduced rank analysis of time varying cointegration vector
Speaker: Xiaoqi Dong
Title: Weighted reduced rank analysis of time varying cointegration vector
Date/Time: 20 November 2020, Friday, 10:00 - 11:00 AM EST