Dr Hongbin GuoOffice: Science III, 310a
- Nonparametric Modeling
- Machine Learning
- Shape-restricted Regression
- Graph Theory
- Dimension Reduction
Parameter Estimation in Extinction / Recolonization Coalesecent Model with Spatial Information
In the field of phylogenetics, understanding population dynamics from the analysis of molecular and spatial data requires sound statistical modeling. Current approaches assume that populations are naturally partitioned into discrete demes, thereby failing to be relevant in cases where individuals are scattered on a spatial continuum. Dr.Stephane Guindon introduced a new genealogy-based inference framework. We used a stochastic model to describe the coalescent process with spatial information. Bayesian maximum likelihood estimators were calculated using Monte-Carlo simulation.
A New Smooth Nonparametric Estimator on Shape-restricted Regression
Estimation of a function under shape restriction is of considerable interest in many practical applications. It is not uncommon that in many fields, researchers are in the position of having strong presumptions about certain relationships satisfying qualitative restrictions, such as monotonicity and convexity (concavity). Typical examples include the study of utility functions, cost functions, and profit functions in economics, the study of dose response curve in medicine, growth curves of animals and plants in ecology, and the estimation of the hazard rate in survival analysis. Imposing shape restrictions can improve the predictive performance and reduce overfitting if the underlying regression function takes the specific form. The classic least squares solutions for shape-restricted estimation are typically neither smooth nor parsimonious. There has been many researches pursuing smooth shape-restricted regressors in recent years.
In our work, we propose two new nonparametric estimators for univariate regression subject to monotonicity and convexity constraints with simple structures, which are acquired by replacing the discrete measures in the non-smooth least squares solutions with continuous ones. Our estimators are composed as the linear combinations of some well-constructed component functions which satisfy corresponding shape constraints. The smoothness of our models are controlled by the tuning parameter. A fast gradient-based iterative algorithm is used to find the least squares estimates with efficiency. Finite sample properties and asymptotic behaviors including the existence, the uniqueness, the equivalence, and the consistency have been investigated. Numerical studies with simulated and real-world data show that our estimators have better prediction performance comparing to other shape-restricted estimators in most scenarios. A series of papers are expected to be published in 2020.
Food Delivering Pathway Optimization
The food delivering industry is booming in the current social environment. The traditional shortest path algorithm is not complicated enough to provide an optimal solution of the riders path. We aim to develop a system specifically designed for the food delivery industry. The target to be optimized in our system is a weighted combination of many variables including the restaurant workflow, the rider’s performance, clients’ sensitivity, real-time traffic information, etc. We believe that this target function will enable us to optimize the profits of all aspects to an overall maximum. By establishing a measuring system for restaurants and riders, we will be able to precisely estimate the food making/delivering time, therefore minimize the time wasted during the “order-cook-pick-deliver” process accordingly. Our system also enables the rider to integrate multiple missions to further improve efficiency. In addition, we creatively designed an algorithm to collect and measure customers’ reviews to assess their sensitivity to different delivery qualities. Customers with relatively higher priorities and sensitivities will be emphasized to escalate the general evaluation of the service provider.
- Demographic inference under the coalescent in a spatial continuum. Stéphane Guindon, Hongbin Guo, David Welch. Theoretical Population Biology, Volume 111, Pages 43-50, 2016.
- 2017 IASC-NZSA conference, Auckland, New Zealand. Oral Presentation.
- Scaffold protein GhMORG1 enhances the resistance of cotton to Fusarium oxysporum by facilitating the MKK6- MPK4 cascade. Wang, Chen; Guo, Hongbin; He, Xiaowen; Zhang, Shuxin; Wang, Jiayu; et al. Plant biotechnology journal, 2019.
- 2020 International Conference of Mathematics and Statistics, Sharjah, UAE. Oral Presentation.
- Analyses of the function of DnaJ family proteins reveal an underlying regulatory mechanism of heat tolerance in honeybee. Guilin Li, Hang Zhao, Hongbin Guo, Ying Wang, Xuepei Cui, Han Li, Baohua Xu, Xingqi Guo. Science of the Total Environment. Volume 716, 10 May 2020.
- Functional and transcriptomic analyses of the NF-Y family provide insights into the defense mechanisms of honeybees under adverse circumstances. Li, G., Zhao, H., Guo, H., Wang, Y., Cui, X., Xu, B., Guo, X. Cellular and Molecular Life Sciences. DOI: 10.1007/s00018-019-03447-0