Theses and Dissertations from UMD
Permanent URI for this communityhttp://hdl.handle.net/1903/2
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
2 results
Search Results
Item FUNCTIONAL PRINCIPAL COMPONENT ANALYSIS WITH APPLICATION TO VIEWERSHIP OF MOTION PICTURES(2014) Tian, Yue; Smith, Paul; Mathematical Statistics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Principal Component Analysis (PCA) is one widely used data processing technique in application, especially for dimensionality reduction. Functional Principal Component Analysis (fPCA) is a generalization of ordinary PCA, which focuses on a sample of functional observations and projects the original functional curves to a new space of orthogonal dimensions to capture the primary features of original functional curves. While, fPCA suffers from two potential error sources. One error source is originated from truncation when we approximate the functional subject's expansion; The other stems from estimation when we estimate the principal components from the sample. We first introduce a generalized functional linear regression model and propose it in the Quasi-likelihood setting. Asymptotic inference of the proposed functional regression model is developed. We also utilize the proposed model to help marketing operational decision process by analyzing viewership of motion pictures. We start with discussing customer reviews effect on movie box office sales. We use the functional regression model with function interactions to measure the effect of Word-of-Mouth on movie box office sales. One main challenge of modeling with functional interactions is the interpretation of model estimate results. We demonstrate one method to help us get important insights from model results by plotting and controlling a re-labbeld 3-D plot. Apart from movie performance in theater, we also employ functional regression model to predict movie pre-release demand in Video-on-Demand (VOD) channel. As its growing popularity, VOD market attracts much attention in marketing research. We analyze the prediction accuracy of our proposed functional regression model with spatial components and find that our proposed model gives us the best predictive accuracy. In summary, the dissertation develops asymptotic properties of a generalized functional linear regression model, and applies the proposed model in analyzing viewership of motion picture both in theater and Video-on-Demand channels. The proposed model not only advances our understanding of motion picture demand, but also helps optimize business decision making process.Item Productivity Dispersion, Plant Size, and Market Structure(2008-06-16) Bakhtiari, Sasan; Haltiwanger, John C; Economics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Ample evidence from micro data suggests that productivity at establishment level is dominated by idiosyncratic factors. The productivity differences across establishments are very large and persistent even with the narrowest definition of industries. There is an attempt to identify sources of frictions that cause such productivity dispersion and negatively affect the average productivity of industries. This dissertation contemplates a non-monotonic relationship between productivity and input size and studies its importance in shaping the relationship between productivity dispersion and the producer size, a fact that is presented along with supportive empirical results. The role of market structure is then elaborated in creating the observed behavior. The US Census of Manufactures reveals significant productivity dispersion at any employment level. Moreover, this productivity dispersion falls with employment size within most manufacturing industries. This pattern is considerably strong for establishments in industries whose products are primarily locally traded. It will be shown that general approaches such as industry selection and simple statistical aggregation do not explain this pattern convincingly, while sector-specific factors such as market localization can mimic this behavior much more closely. Based on these results, a market structure model is introduced that uses demand size and market localization as constraining forces to generate a bell-shaped relationship between input size and productivity within a market and for locally traded goods. The non-monotonicity of the relationship is a clear departure from most economic models where input size of plants is monotonically increasing with their productivity in the long-run. Because of the bell-shaped relationship, the proposed model predicts significant long-run productivity dispersion at any level of input size. Also this dispersion decreases with input size, in the same way as is observed in the data. The model is calibrated and then simulated using data on Ready-Mix Concrete. First, the relationship between productivity and input size in the data is of a similar bell-shaped form. The effect of market size is also shown to be consistent with model predictions. Second, simulated results produce productivity dispersions that fall with input size with almost the same slope as observed in the data. This, in turn, suggests that the difference between simulated and actual productivity dispersions, summarizing the effect of other frictions, is almost uniform across sizes. Finally the robustness of the results is demonstrated through various tests. Throughout the discussion, a distinction is made between physical and revenue productivities and the theoretical implications of both measures are shown to be qualitatively the same.