A5102718627- Stochastic processes and financial applications
- Statistical Methods and Inference
- Financial Risk and Volatility Modeling
- Point processes and geometric inequalities
- Random Matrices and Applications
- Probability and Risk Models
- Monetary Policy and Economic Impact
- Mathematical Approximation and Integration
- Bayesian Methods and Mixture Models
- Markov Chains and Monte Carlo Methods
- Insurance, Mortality, Demography, Risk Management
- advanced mathematical theories
- Risk and Portfolio Optimization
- Domain Adaptation and Few-Shot Learning
- Stochastic processes and statistical mechanics
- Statistical Methods and Bayesian Inference
- Advanced Statistical Methods and Models
- Cognitive Computing and Networks
- Graph theory and applications
- Advanced Mathematical Theories and Applications
- Data Stream Mining Techniques
- Complex Network Analysis Techniques
- Artificial Intelligence in Games
- Spatial and Panel Data Analysis
- Quantum Computing Algorithms and Architecture
Shandong University
2014-2023
Shandong University of Finance and Economics
2022
China Institute of Finance and Capital Markets
2014-2020
Tianjin University
2020
Zhejiang University of Finance and Economics
2014-2020
Zhejiang University
2010-2015
University of Science and Technology of China
2014
Transfer learning through fine-tuning a pre-trained neural network with an extremely large dataset, such as ImageNet, can significantly accelerate training while the accuracy is frequently bottlenecked by limited dataset size of new target task. To solve problem, some regularization methods, constraining outer layer weights using starting point references (SPAR), have been studied. In this paper, we propose novel regularized transfer framework DELTA, namely DEep Learning Feature Map...
In this paper, we study the asymptotic error distribution for a two-level irregular discretization scheme of solution to stochastic differential equations (SDE short) driven by continuous semimartingale and obtain central limit theorem processes with rate $\sqrt{n}$. As an application, in spirit result Ben Alaya Kebaier, get Linderberg-Feller type multilevel Monte Carlo method.
In this paper, we consider estimating spot/instantaneous volatility matrices of high-frequency data collected for a large number assets. We first combine classic nonparametric kernel-based smoothing with generalized shrinkage technique in the matrix estimation noise-free under uniform sparsity assumption, natural extension approximate commonly used literature. The consistency property is derived proposed spot estimator convergence rates comparable to optimal minimax one. For contaminated by...
Limit theory involving stochastic integrals is now widespread in time series econometrics and relies on a few key results functional weak convergence. In establishing such convergence, the literature commonly uses martingale semimartingale structures. While these structures have wide relevance, many applications involve cointegration framework where endogeneity nonlinearity play major roles complicate limit theory. This paper explores convergence to integral functionals settings. We use...
Abstract In this article, we present the local linear estimations for diffusion coefficient and drift in second-order model. We show that under mild conditions, estimators are weak consistent. also use a Monte Carlo experiment to compare our with ones Nicolau (2007 , J. ( 2007 ). Nonparametric estimation of scend-order stochastic differential equations . Econometric Theor. 23 : 880 – 898 .[Crossref], [Web Science ®] [Google Scholar]). Keywords: Local estimationN-W estimatorSecond-order...
In this paper we propose the asymptotic error distributions of Euler scheme for a stochastic differential equation driven by Itô semimartingales. Jacod (2004) studied problem equations pure jump Lévy processes and obtained quite sharp results. We extend his results to more general semimartingale.
Abstract Central limit theorems play an important role in the study of statistical inference for stochastic processes. However, when non‐parametric local polynomial threshold estimator, especially linear case, is employed to estimate diffusion coefficients processes, adaptive and predictable structure estimator conditionally on σ ‐field generated by processes destroyed, so classical central theorem martingale difference sequences cannot work. In high‐frequency data, we proved estimators...
War-game is a type of multi-agent real-time strategy game, with challenges the large-scale decision-making space and flexible changeable battlefield situation. In addition to military field, it has played role in fields including epidemic prevention pest control. recent years, more learning algorithms have tried solve this kind game. However, existing methods not yet given satisfactory solution for war-game, especially when preparation time limited. background, we try traditional war-game...
Weak convergence of various general functionals partial sums dependent random variables to stochastic integral now play a major role in the modern statistics theory. In this paper, we obtain weak casual process by means method which was introduced Jacod and Shiryaev (2003).
This paper studies estimation of covariance matrices with conditional sparse structure. We overcome the challenge estimating dense using a factor structure, large-dimensional by postulating sparsity on random noises, and varying allowing loadings to smoothly change. A kernel-weighted approach combined generalised shrinkage is proposed. Under mild conditions, we derive uniform consistency for developed method obtain convergence rates. Numerical including simulation an empirical application...
We study the nonparametric estimators of infinitesimal coefficients second-order jump-diffusion models. Under mild conditions, we obtain weak consistency and asymptotic normalities estimators.
Limit theory involving stochastic integrals is now widespread in time series econometrics and relies on a few key results function space weak convergence. In establishing convergence of sample covariances to integrals, the literature commonly uses martingale semimartingale structures. While these structures have wide relevance, many applications involve cointegration framework where endogeneity nonlinearity play major role lead complications limit theory. This paper explores integral...
In this paper we propose the asymptotic error distributions of Euler scheme for a stochastic differential equation driven by Itô semimartingales. Jacod (2004) studied problem equations pure jump Lévy processes and obtained quite sharp results. We extend his results to more general semimartingale.
With the rapid development of classical and quantum machine learning, a large number learning frameworks have been proposed. However, existing usually only focus on or quantum, rather than both. Therefore, based VQNet 1.0, we further propose 2.0, new generation unified framework that supports hybrid optimization. The core library is implemented in C++, user level Python, it deployment hardware. In this article, analyze trend introduce design principles 2.0 detail: unity, practicality,...
In this paper, we consider estimating spot/instantaneous volatility matrices of high-frequency data collected for a large number assets. We first combine classic nonparametric kernel-based smoothing with generalised shrinkage technique in the matrix estimation noise-free under uniform sparsity assumption, natural extension approximate commonly used literature. The consistency property is derived proposed spot estimator convergence rates comparable to optimal minimax one. For contaminated by...
In the study of supremum stochastic processes, Talagrand's chaining functionals and his generic method are heavily related to distribution processes. present paper, we construct type in general case obtain upper bound for suprema all $p$-th moments process using method. As applications, obtained Johnson-Lindenstrauss lemma, moment order 2 Gaussian chaos, convex signal recovery our setting.
We develop a novel decoupling inequality for nonhomogeneous order $2$ chaos process. Having this inequality, we show bound the suprema of log-concave tailed process, based on some chaining techniques and generalized majorizing measure theorems canonical processes. As applications, R.I.P. partial random circulant matrices time-frequency structured induced by standard $\alpha$-subexponential vectors, $1\le \alpha\le 2$, which extends previously known results subgaussian case.
Let $M$ be an $n\times n$ random i.i.d. matrix. This paper studies the deviation inequality of $s_{n-k+1}(M)$, $k$-th smallest singular value $M$. In particular, when entries are subgaussian, we show that for any $\gamma\in (0, 1/2), \varepsilon>0$ and $\log n\le k\le c\sqrt{n}$ \begin{align} \textsf{P}\{s_{n-k+1}(M)\le \frac{\varepsilon}{\sqrt{n}} \}\le \Big( \frac{C\varepsilon}{k}\Big)^{\gamma k^{2}}+e^{-c_{1}kn}.\nonumber \end{align} result improves existing Nguyen, which obtained a...