School of Mathematics and Statistics - Research Publications
Permanent URI for this collection
Search Results
1 - 8 of 8
-
ItemNo Preview AvailableGenerating Dynamic Kernels via Transformers for Lane Detection(IEEE, 2023-01-01)
-
ItemPinpointing Early Signs of Impending Slope Failures From Space(AMER GEOPHYSICAL UNION, 2022-02)Abstract A promised potential of spaceborne interferometric synthetic aperture radars (InSAR) is a capability for regularly monitoring ground deformation with millimeter accuracy, for timely forecasting of impending natural hazards such as landslides. The main limitation in InSAR being actually capable of unleashing this potential for hazard prediction is that key precursory ground displacements are, in the majority of cases, a very small subset of the entire big data set provided by the method over large regions. Consequently, pinpointing a single impending failure may become very difficult or impossible. We develop a data‐driven framework that can handle such imbalanced spatiotemporal data based on the concept of outlying aspects mining, to find a subset of features out of a collection of potential features, which best distinguishes the landslide source area from the others. We show that the identified feature subspace can be used to find anomalous areas across multiple spatial scales, such that Sentinel‐1 satellite monitoring points which persistently lie in these areas can accurately detect the location of the Xinmo landslide (China) almost 1 year in advance—without false alarms. In a second case study, we identify the area affected by rockfalls on Stromboli volcano, a task that is generally infeasible with traditional methods applied to InSAR data. With continuing improvements in the spatial and temporal resolution from the new generation of satellites, such as Sentinel‐1, this approach opens the door to reliable and early prediction of failure over a broad range of slope instabilities.
-
ItemNo Preview AvailableThe Prediction for COVID-19 Outbreak in China by using the Concept of Term Structure for the Turning Period(ELSEVIER SCIENCE BV, 2021)This study aims to develop a general framework for predicting the duration of the Turning Period (or Turning Phase) for the COVID-19 outbreak in China that started in late December 2019 from Wuhan. A new concept called the Term Structure for Turning Period (instead of Turning Point) is used for this study, and the framework, implemented into an individual SEIR (iSEIR) model, has enabled a timely prediction of the turning period when applied to Wuhan's COVID-19 epidemic, and provided the opportunity for relevant authorities to take appropriate and timely actions to successfully control the epidemic. By using the observed daily COVID-19 cases in Wuhan from January 23, 2020 to February 6 (and February 10), 2020 as inputs to the framework it allowed us to generate the trajectory of COVID-19 dynamics and to predict that the Turning Period of COVID-19 outbreak in Wuhan would arrive within one week after February 14. This prediction turned out to be timely and accurate, which has provided adequate time for the government, hospitals and related sectors and services to meet peak demand and to prepare aftermath planning. We want to emphasize that emergency risk management entails the implementation of an emergency plan, where timing the Turning Period is key to express a clear timeline for effective actions. Our study confirms the observed effectiveness of Wuhan's Lockdown and Isolation control program imposed since January 23, 2020 to the middle of March, 2020 and resulted in swiftly flattened epidemic curve, and Wuhan's success offers an exemplary lesson for the world to learn in combating COVID-19 pandemic.
-
ItemAssociation between the yield and the main agronomic traits of Tartary buckwheat evaluated using the random forest model(ACSESS, 2020-09-01)The popularity of Tartary buckwheat [Fagopyrum tataricum (L.) Gaertn], as a medicinal and food crop, has been increasing in recent years. However, its low yield seriously restricts its industrial development. Amongst the various studies conducted to enhance the productivity of Tartary buckwheat, the association between yield and main agronomic traits has formed the foundation for the breeding and cultivation of high-yielding varieties, becoming the primary interest of breeders. The commonly used methods are often restricted by sample size, distribution assumptions and trait properties and confined to the linear relationship. In this paper, the random forest regression model was used to obtain a comprehensive and reliable evaluation. The phenotypic data of 200 Tartary buckwheat landraces with 15 quantitative and two qualitative agronomic traits for two consecutive years were used. Results were compared between planting seasons and with those from classical methods, such as the correlation analyses and the multiple linear regression model. The random forest model distinguished the number of grains per plant, plant height, and 1,000-grain weight as the most influential agronomic traits in both seasons. The main and interactive effects were explored using the accumulated local effects plot and showed great conformity between the two seasons. The robustness and reliability of the random forest model make it a desirable methodology for breeding new varieties and germplasm innovation of Tartary buckwheat and other crops.
-
ItemSemisupervised Clustering by Iterative Partition and Regression with Neuroscience Applications(HINDAWI LTD, 2016)Regression clustering is a mixture of unsupervised and supervised statistical learning and data mining method which is found in a wide range of applications including artificial intelligence and neuroscience. It performs unsupervised learning when it clusters the data according to their respective unobserved regression hyperplanes. The method also performs supervised learning when it fits regression hyperplanes to the corresponding data clusters. Applying regression clustering in practice requires means of determining the underlying number of clusters in the data, finding the cluster label of each data point, and estimating the regression coefficients of the model. In this paper, we review the estimation and selection issues in regression clustering with regard to the least squares and robust statistical methods. We also provide a model selection based technique to determine the number of regression clusters underlying the data. We further develop a computing procedure for regression clustering estimation and selection. Finally, simulation studies are presented for assessing the procedure, together with analyzing a real data set on RGB cell marking in neuroscience to illustrate and interpret the method.
-
Item
-
ItemSetwise and Filtered Gibbs Samplers for Teletraffic Analysis(ASSOC COMPUTING MACHINERY, 2010-04)A setwise Gibbs sampler (SGS) method is developed to simulate stationary distributions and performance measures of network occupancy of Baskett-Chandy-Muntz-Palacios (BCMP) telecommunication models. It overcomes the simulation difficulty encountered in applying the standard Gibbs sampler to closed BCMP networks with constant occupancy constraints. We show Markov chains induced by SGS converge to the target stationary distributions. This article also investigates the filtered Gibbs sampler (FGS) as an efficient method for estimating various network performance measures. It shows that FGS's efficiency is considerable, but may be improperly overestimated. A more conservative performance estimator is then presented.
-
ItemMODEL SELECTION AND CLAIM FREQUENCY FOR WORKERS' COMPENSATION INSURANCE(CAMBRIDGE UNIV PRESS, 2010-11)