Wen Cheng

ORCID: 0000-0003-3082-6746
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Stochastic processes and financial applications
  • Differential Equations and Numerical Methods
  • Time Series Analysis and Forecasting
  • Anomaly Detection Techniques and Applications
  • Neural Networks and Applications
  • Stock Market Forecasting Methods
  • Advanced Mathematical Modeling in Engineering
  • Advanced Algorithms and Applications
  • advanced mathematical theories
  • Financial Risk and Volatility Modeling
  • Advanced Computational Techniques and Applications
  • Forecasting Techniques and Applications
  • Target Tracking and Data Fusion in Sensor Networks
  • Complex Systems and Time Series Analysis
  • Music and Audio Processing
  • Energy Load and Power Forecasting
  • Fault Detection and Control Systems
  • Financial Markets and Investment Strategies
  • Functional Brain Connectivity Studies
  • Machine Learning and ELM
  • Advanced Decision-Making Techniques
  • Differential Equations and Boundary Problems
  • Emotion and Mood Recognition
  • Infrared Target Detection Methodologies
  • Stability and Controllability of Differential Equations

Bengbu Medical College
2024

Nanyang Technological University
2017-2023

Pennsylvania State University
2009-2014

Nanjing University of Aeronautics and Astronautics
2013

Xi'an High Tech University
2009

Copiah-Lincoln Community College
2001

10.1016/j.engappai.2022.105349 article EN Engineering Applications of Artificial Intelligence 2022-09-05

Hospitals can predetermine the admission rate and facilitate resource allocation based on valid emergency requests bed capacity estimation. The excess unoccupied beds be determined with help of forecasting number discharged patients. Extracting predictive features mining temporal patterns from historical observations are crucial for accurate reliable forecasts. Machine learning algorithms have demonstrated ability to learn knowledge make predictions unseen inputs. This paper utilizes several...

10.1109/jbhi.2022.3172956 article EN IEEE Journal of Biomedical and Health Informatics 2022-05-06

We construct closed-form asymptotic formulas for the Green's function of parabolic equations (e.g., Fokker–Planck equations) with variable coefficients in one space dimension. More precisely, let $u(t,x)=\int\mathcal{G}_t(x,y)f(y)dy$ be solution $\partial_tu-(au"+bu'+cu)=0$ $t>0$, $u(0,x)=f(x)$. Then we find computable approximations $\mathcal{G}_t^{[n]}$ $\mathcal{G}_t$. The approximate kernels are derived by applying Dyson–Taylor commutator method that have recently developed short-time...

10.1137/100796832 article EN SIAM Journal on Financial Mathematics 2011-01-01

Heterogeneous findings among anxiety disorder studies have hindered elucidation of the underlying pathophysiology and development mechanism-based therapies. To determine whether structural MRI in converge on a common network with therapeutic significance. In this retrospective study, systematic literature search PubMed Web Science databases was performed to identify coordinates gray matter atrophy patients disorder. Atrophy were then mapped an constructed from resting-state functional...

10.1155/2024/3827870 article EN Depression and Anxiety 2024-04-02

This paper proposes a performance-driven gradient boosting model (pdGBM) which predicts short-horizon price movements by combining nonlinear response functions of selected predictors. performs descent in constrained functional space directly minimizing loss customized with different trading performance measurements. To demonstrate its practical applications, simple system was designed signals constructed from pdGBM predictions and fixed holding period each trade. We tested this on the...

10.1080/14697688.2015.1032541 article EN Quantitative Finance 2015-07-09

Decision Tree is a simple but popular machine learning algorithm. Although single decision tree not as accurate other state-of-the-art classifiers, the performance can be significantly improved by combining predictions of several trees i.e. creating an ensemble trees. In this paper, we study and their ensembles viz. Bagged Trees, Random Forest, Extremely Randomized Rotation Gradient Boosted Trees AdaBoosted assess on UCI datasets. addition, propose new method, Heterogeneous Ensemble trees,...

10.1109/ssci.2017.8285445 article EN 2021 IEEE Symposium Series on Computational Intelligence (SSCI) 2017-11-01

We obtain new closed-form pricing formulas for contingent claims when the asset follows a Dupire-type local volatility model. To we use Dyson-Taylor commutator method recently developed in [7, 8, 10] short time asymptotic expansions of heat kernels, and family general explicit closed form approximate solutions both kernel derivative price. also perform analytic as well numerical error analysis, compare our results to other known methods.

10.2139/ssrn.1486470 article EN SSRN Electronic Journal 2009-01-01

Inspired by recent work on the Dyson-Taylor Commutator method, we apply similar but different techniques to commodity spread options when correlation (denoted Ρ) of two underlying Brownian drivers is close 1. We obtain closed-form asymptotic formulas in powers 1−Ρ.

10.2139/ssrn.2533310 article EN SSRN Electronic Journal 2014-01-01

A novel object tracking algorithm for FLIR imagery based on mean shift using multiple features is proposed to improve the performance. First, appearance model of infrared represented in combination gray space, LBP texture and orientation space with different feature weight. And then, employed find location. An on-line weight update mechanism developed Fisher criteria, which measure discrimination background effectively. Experiment results demonstrate effectiveness robustness method imagery.

10.1117/12.832386 article EN Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE 2009-09-29

We obtain new closed-form pricing formulas for contingent claims when the asset follows a Dupire-type local volatility model. To we use Dyson-Taylor commutator method that have recently developed in [5, 6, 8] short-time asymptotic expansions of heat kernels, and family general approximate solutions both kernel derivative price. A bootstrap scheme allows us to extend our large time. also perform analytic as well numerical error analysis, compare results other known methods.

10.48550/arxiv.0910.2309 preprint EN other-oa arXiv (Cornell University) 2009-01-01

This note introduces a mathematically rigorous short time approximation of the transition density function CEV model. We first apply change variable to operator and transform it Schrodinger with an inverse square potential, then construct Neumann series new opeator under weighted Sobolev spaces. To author’s knowledge, this is that construction for model obtained in financial mathematics literature.

10.2139/ssrn.2487669 article EN SSRN Electronic Journal 2014-01-01
Coming Soon ...