Browsing by Subject "Time series analysis"
Now showing 1 - 5 of 5
Results Per Page
Sort Options
Item A comparison of seismic site response methods(2010-08) Kottke, Albert Richard; Rathje, Ellen M.; Gilbert, Robert B.; Stokoe, III, Kenneth H.; Manuel, Lance; Grand, Stephen P.Local soil conditions influence the characteristics of earthquake ground shaking and these effects must be taken into account when specifying ground shaking levels for seismic design. These effects are quantified via site response analysis, which involves the propagation of earthquake motions from the base rock through the overlying soil layers to the ground surface. Site response analysis provides surface acceleration-time series, surface acceleration response spectra, and/or spectral amplification factors based on the dynamic response of the local soil conditions. This dissertation investigates and compares the results from different site response methods. Specifically, equivalent-linear time series analysis, equivalent-linear random vibration theory analysis, and nonlinear time series analysis are considered. In the first portion of this study, hypothetical sites and events are used to compare the various site response methods. The use of hypothetical events at hypothetical sites allowed for the seismic evaluation process used in engineering practice to be mimicked. The hypothetical sites were modeled after sites with characteristics that are representative of sites in the Eastern and Western United States. The input motions selected to represent the hypothetical events were developed using the following methods: stochastically-simulated time series, linearly-scaled recorded time series, and spectrally-matched time series. The random vibration theory input motions were defined using: seismological source theory, averaging of the Fourier amplitude spectra computed from scaled time series, and a response spectrum compatible motion. All of the different input motions were then scaled to varying intensity levels and propagated through the sites to evaluate the relative differences between the methods and explain the differences. Data recorded from borehole arrays, which consist of instrumentation at surface and at depth within the soil deposit, are used to evaluate the absolute bias of the site response methods in the second portion of this study. Borehole array data is extremely useful as it captures both the input motion and the surface motion, and can be used to study solely the wave propagation process within the soil deposit. However, comparisons using the borehole data are complicated by the assumed wavefield at the base of the array. In this study, sites are selected based on site conditions and the availability of high intensity input motions. The site characteristics are then developed based on site specific information and data from laboratory soil testing. Comparisons between the observed and computed response are used to first assess the wavefield at the base of the array, and then to evaluate the accuracy of the site response methods.Item Dirty statistical models(2012-05) Jalali, Ali, 1982-; Sanghavi, Sujay Rajendra, 1979-; Caramanis, Constantine; Ghosh, Joydeep; Dhillon, Inderjit; Ravikumar, PradeepIn fields across science and engineering, we are increasingly faced with problems where the number of variables or features we need to estimate is much larger than the number of observations. Under such high-dimensional scaling, for any hope of statistically consistent estimation, it becomes vital to leverage any potential structure in the problem such as sparsity, low-rank structure or block sparsity. However, data may deviate significantly from any one such statistical model. The motivation of this thesis is: can we simultaneously leverage more than one such statistical structural model, to obtain consistency in a larger number of problems, and with fewer samples, than can be obtained by single models? Our approach involves combining via simple linear superposition, a technique we term dirty models. The idea is very simple: while any one structure might not capture the data, a superposition of structural classes might. Dirty models thus searches for a parameter that can be decomposed into a number of simpler structures such as (a) sparse plus block-sparse, (b) sparse plus low-rank and (c) low-rank plus block-sparse. In this thesis, we propose dirty model based algorithms for different problems such as multi-task learning, graph clustering and time-series analysis with latent factors. We analyze these algorithms in terms of the number of observations we need to estimate the variables. These algorithms are based on convex optimization and sometimes they are relatively slow. We provide a class of low-complexity greedy algorithms that not only can solve these optimizations faster, but also guarantee the solution. Other than theoretical results, in each case, we provide experimental results to illustrate the power of dirty models.Item International black tea market integration and price discovery(Texas A&M University, 2004-09-30) Dharmasena, Kalu Arachchillage Senarath Dhananjaya BandaraIn this thesis we study three basic issues related to international black tea markets: Are black tea markets integrated? Where is the price of black tea discovered? Are there leaders and followers in black tea markets? We use two statistical techniques as engines of analysis. First, we use time series methods to capture regularities in time lags among price series. Second, we use directed acyclic graphs to discover how surprises (innovations) in prices from each market are communicated to other markets in contemporaneous time. Weekly time series data on black tea prices from seven markets around the world are studied using time series methods. The study follows two paths. We study these prices in a common currency, the US dollar. We also study prices in each country's local currency. Results from unit root tests suggest that prices from three Indian markets are not generated through random walk-like behavior. We conclude that the Indian markets are not weak form efficient. However, prices from all non-Indian markets cannot be distinguished from random walk-like behavior. These latter markets are weak form efficient. Further analysis on these latter markets is conducted to determine whether information among the markets is shared. Vector Autoregressions (VARs) on the non-Indian markets are studied using directed acyclic graphs, impulse response functions and forecast error decomposition analyses. In both local currencies and dollar-converted series, the Sri Lankan and Indonesian markets are price leaders in contemporaneous time. Kenya is an information sink. It is endogenous in current time. Malawi is an exogenous price leader in dollar terms, but it is endogenous in local currency in contemporaneous time. In the long run, Sri Lanka, Indonesia and Malawi are price leaders in US dollar terms. In local currency series, Indonesia, Kenya and Malawi are price leaders in the long run. We use Theil's U-statistic to test the forecasting ability of the VAR models. We find for most markets in either dollars or on local currencies that a random walk forecast outperforms the VAR generated forecasts. This last result suggests the non-Indian markets are both weak form and semi-strong form efficient.Item Time series models of the electrical conductivity measured at the Manchar Lake in Pakistan(2010-05) Zehra, Syeda Mahe; Barnes, J. Wesley; Pierce, Suzanne A.The Manchar Lake in Pakistan is in danger. So are the native fisher folk that populate the area and lake. The lake is undergoing water quality degradation due to both a decrease in the amount of freshwater inflow and an increase in the polluted agricultural run-off. The fish in the lake are dying and some fish species are becoming extinct, the people living on and around the lake are facing severe health risks, migratory birds are no longer stopping at the Manchar Lake and agriculture in the area is also suffering. This report focuses on time-series modeling and analysis of water quality data from Manchar Lake. We evaluate data for three sites within the Manchar Lake and complete time series models and analysis for two sites. Particular attention is given to the Electrical Conductivity data of the lake. The approach to modeling and time series analysis provide an example of potential uses of measured data to recognize shifts in water quality within the context of potential insights and recommendations about lake management in the area.Item Water policy informatics : a topic and time series analysis of the Texas state water plans(2011-05) Wehner, Jenifer Elizabeth; Pierce, Suzanne Alise, 1969-The disciplines of informatics and information visualization have developed in response to societal needs to find new insight in complex datasets and have been enabled by technological advancements. Joint application of these fields can demonstrate themes and connections that are otherwise not apparent. Methodological approaches, such as direct network analysis, can be applied to policy documents to determine if action or policy recommendations match the goals or objectives stated in the within the same documents. Informatics and information visualization can also be used to analyze changes of themes found within the documents over time. This paper seeks to leverage informatics and information visualization methodologies as a novel approach to policy analysis. In particular, directed network and time burst techniques are used to analyze water management policy documents for the State of Texas. The congruency between the stated goals or objectives and recommendations sections is evaluated at a topical level within each planning document and possible changes in important water policy concepts over time are highlighted by comparing among multiple planning documents. Although there limitations to the process at the time of publication due to the newness of the software utilized, this paper demonstrates that the products still lead to unique and insightful conclusions.