363 research outputs found
Bayesian high dimensional modeling with group structures
Full text linkU of I OnlySubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2022-05-01The student, Xinming Yang, accepted the attached license on 2020-05-05 at 20:34.The student, Xinming Yang, submitted this Dissertation for approval on 2020-05-05 at 20:54.This Dissertation was approved for publication on 2020-05-06 at 18:17.DSpace SAF Submission Ingestion Package generated from Vireo submission #15226 on 2020-08-25 at 17:29:47Made available in DSpace on 2020-08-26T23:58:39Z (GMT). No. of bitstreams: 2
YANG-DISSERTATION-2020.pdf: 3544454 bytes, checksum: e3aa183dd42b7e93ee98fcce7520d4ee (MD5)
LICENSE.txt: 4209 bytes, checksum: d42ae85ec45ea6f48b277cd30ef4b831 (MD5)
Previous issue date: 2020-05-06Embargo set by: Seth Robbins for item 115771
Lift date: 2022-08-26T23:58:55Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemGroup structures arise naturally in a variety of modern data applications and statistical problems in the high-dimensional data setting where the number of variables can greatly exceed the number of observations. The group structures are usually very informative as they express the inherent similarities among the variables and observations and it is thus desirable to take the prior group information into consideration in the construction of statistical models in pursuit of efficient statistical inference. In this dissertation, we propose methods for three statistical problems: linear regression, graphical model, and sequential logistic regressions when group structures are present at either the variable level or the observation level. We adopt the Bayesian framework and extend the spike-and-slab priors in the group setting to incorporate the group information. Our proposed hierarchical Bayesian models are well suited for sharing of similar sparsity patterns within the groups. For posterior computation, we propose EM algorithms and shotgun stochastic search algorithm which are more efficient than standard Markov chain Monte Carlo sampling algorithms. Compared to methods that simply ignore the grouping structure, our proposed methods that involve the group information can lead to better statistical inference results and the fitted models are more interpretable and provide more insights into the data. Further, we show that our methods are also advantageous theoretically or empirically in comparison with other group selection competitors due to the nonconvex regularization induced from our Bayesian modeling.Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD syste
Real-Time Hybrid In-Station Bus Dispatching Strategy Based on Mixed Integer Programming
No full textThe actual bus headway often deviates from the planned departure frequency because of external factors, such as traffic conditions and public transport demand, leading to transit resource waste and reducing the quality of service. In view of the existing shortcomings of the current dispatching strategy, a mixed integer programming model, integrating a bus-holding and stop-skipping strategy, is constructed to improve transit service with a minimum cost. The real-time optimal holding and stop-skipping strategies can be obtained by solving the proposed model using the Lagrangian relaxation algorithm. A numerical example is conducted using real transit GPS (Global Position System) and IC (Intelligent Card) data in Harbin. The results show that compared to a single control strategy, the proposed hybrid model is a better trade-off between the quality of the transit service and the operation cost. Notably, such a strategy would produce a minimal passengers’ average travel time coefficient. It is a great help for promoting the transit service level and increasing competitiveness
- …
