The problem considered in this study arose from the need for an emperlcal analysis of the sale value of land to determine if the increased price benefits of the governmental flue-cured tobacco program have been absorbed by higher land rents. If acreage allotments giving the right to produce tobacco under the program are capitalized into farmland values to an appreciable extent, the program objective of increasing farm incomes would be partially defeated through higher rents. The objective of this study was to determine the extent to which allotments have been capitalized into land values. Data on sale value of farms and factors expected to influence the farm sale value were secured from primary public record sources for the four-year period from 1954 to 1957 in two distinctly different flue-cured tobacco regions--Pittsylvania County, Virginia, and Wilson, Greene, and Pitt Counties, North Carolina. These data were analyzed by a multiple regression statistical technique designed to measure the value of an acre of tobacco allotment es a right to produce. The statistical coefficients indicated that an acre of tobacco allotment increased in value from $962 in 1954 to $1,673 in 1957 for Pittsylvania County and from $1,830 to $3,308 for Wilson, Greene, and Pitt Counties. The size of the values for an acre of tobacco allotment as well as the increase in values over the four-year period during which allotments were reduced under the program by 33 percent indicate that an appreciable proportion of the price·rais1ng benefits of the program have been capitalized into land values. / Master of Science

Gender equality is not only a fundamental human right, but a necessary foundation for a peaceful, prosperous and sustainable world. Whilst there has been significant progress as part of a fundamental societal shift, a report by the UN (2022) has identified a further widening of the gender poverty gap. This study takes the perspective of evaluating property valuation of the matrimonial home in divorce cases and how this may contribute to gender inequality. Combining the areas of property valuation and gender inequality is a unique aspect of this study. Currently, this is an under researched area in real estate and family law and this thesis aims to fill this gap. The thesis includes a literature review, which provides a critical assessment of valuation evidence as a data source in research involving a discussion of valuation accuracy, valuation variation and valuation bias. The Australian family court system and the role of valuers as expert witnesses are also discussed in the literature review. Family law and divorce in the context of other countries worldwide are considered as well. The aim of this thesis is to demonstrate through analysis the role of property valuation in family law and in particular retention of the matrimonial home in long-term effects on women following divorce. This can include both financial and non-financial aspects. In order to quantify these aspects, the research in this thesis uses both qualitative and quantitative methods to introduce a set of measurable elements. Data was taken from the Austlii to create a detailed database of 658 cases to be analysed. Quantitative analysis included examination of property valuations put forward to the judge by male and female parties and their effects on final outcomes. Qualitative analysis was conducted with NVivo on a selection of 20 cases from the database in order to identify important themes related to the matrimonial home, valuation, gender roles, socioeconomic factors, and reasoning behind judges’ decisions. Qualitative analysis of a series of interviews with family lawyers, valuers, and clients was also undertaken. Overall findings showed that women are disadvantaged financially following divorce for many reasons, including caretaking duties, limited earning capacity, and career disruptions. There is evidence that judges were aware of such limitations and used these disadvantages as reason to award the women a higher portion of the assets, in particular the matrimonial home. The inference from this evidence does not support the hypothesis of bias against women in the family courts. However, the case studies of women who had been through divorce themselves raised issues that complicate this conclusion. Even when women were awarded the family home, they encountered long-term financial difficulties. From these outcomes, it appears that the strategies used to compensate are not effective in solving the whole problem. Whilst the disadvantage women face has decreased over time, the evidence indicates that there is still scope for improvement.

Initial public offerings (IPO) are an important function for society to fund entrepreneurs. However, what explains the number is disputed. One may think that it should be explained by the need for capital but only a part can be explained by that. Instead, this thesis theories that the number of IPOs should be able to be explained by valuation. To test this hypothesis two variables have been thought to affect the valuation: the current market valuation and underpricing. To represent the current valuation the P/E ratio of OMXS30 was used and for underpricing the average underpricing was used. Which later was tested in a regression analysis against the number of IPOs. This study found that there exists a positive correlation between the current market valuation (P/E) and the number of IPOs. The study's result regarding if underpricing could explain the number of IPOs became inconclusive. The finding of P/E ratios correlation to the number of IPOs can in future forecasting be tested and used.

This dissertation innovatively extends the hedonic value model by crafting a comprehensive and novel econometric framework to analyze the mediating role of financial performance in the valuation of vacation rental properties. The study not only conducts an empirical analysis of the fundamental hedonic determinants of vacation rentals’ performance and value but also examines the influence of vacation rental performance on their valuation by leveraging the capitalization rate theory, a prevalent metric for evaluating commercial real estate value, to ascertain its applicability in forecasting the worth of vacation rentals. The core contribution of this study is the discovery of how short-term rentals are assessed in Central Florida, the key factors contributing to their performance and value, and whether the current valuation methods are suited to the current market needs. The methodology is grounded in robust panel data analysis, encompassing annual performance metrics from the Central Florida market between 2014 and 2023. This approach integrates short-term rentals' extended property-level information and monthly performance databases from AirDNA with the comprehensive real estate database offered by Stellar-MLS, which includes property listing history, market sales, and property details available exclusively to licensed realtors in Central Florida. To integrate these two datasets, a location-based tool was developed using Python and Google API resources to match property addresses and calculate the total distance and driving time to main attractions in the area. Results confirmed the influence of performance on vacation rental valuation and the mediation effect of performance on the relationship between property attributes and value. The findings revealed that the valuation of vacation rentals in Central Florida is better suited to commercial-oriented real estate assessments rather than traditional residential appraisals. These transformative insights provide a robust econometric tool for better predicting the value and performance of vacation rentals, benefiting a range of stakeholders, including vacation rental owners, investors, and policymakers.

This dissertation consists of three essays, of which, two involve assessing the value of tree cover in urban communities and the other evaluates cooperative management of an invasive species by urban communities. The first chapter summarizes the three topics and briefly describes the motivation, methods applied and main conclusions in each study. Chapter 2 presents a meta-analysis of hedonic property value studies on the value of tree cover. A meta-regression was performed using implicit value estimates for tree cover at property-level from various locations in the U.S. along with auxiliary data of county-level tree cover to investigate the relationship between tree cover and implicit-price estimates for residential properties. The study found that on average 35 percent and 40 percent tree cover respectively at property-level and county-level optimize the benefits to the property owners in urban areas. These results provide insights to forward-looking communities to adapt their tree planting and protection efforts to lessen climate-induced impacts. Chapter 3 applies a first-stage Hedonic property price model to estimate preference for tree cover in urban communities using single-family house sales data from multiple property markets across the U.S. The study analyses how home owners' preference for tree cover vary across the landscape and across cities. Further, it identifies what factors affect these variations via the general inferences obtained from an internal meta-analysis. The study confirms the heterogeneity of preferences as affected by the differences in the abundance of tree cover in study locations, regional differences and household characteristics. These findings add to the hedonic literature and provide useful information for future urban planning. Chapter 4 focuses on cooperative management of invasive species in landscapes with mixed land ownerships. This study analyzes the effect of the land ownership on the management efforts between an infested municipality and an uninfested municipality when a transferable payment scheme is involved in the cooperative agreement. A dynamic optimization problem was set up to evaluate the case of Emerald ash borer (EAB) control in multiple jurisdictions in the Twin Cities, Minnesota. The results suggest that when the infested municipality has more public lands and when the transfer payments are efficiently used to implement greater control, the municipalities are more likely to commit to bargaining, and smaller transfer payments paid over a longer span of time are sufficient for optimal control of the spread of invasive species across the municipalities. The last chapter concludes the three studies and discusses the insights for future research. / Ph. D. / Mountain pine beetles (Dendroctonus ponderosae) are a native insect that has decimated millions of hectares of mature pine forests in western North America. The purpose of this study was to investigate, using GIS-derived variables, biophysical and climatic factors that have influenced past mountain pine beetle insect outbreaks, as evident by beetle-induced tree mortality in the Crown of the Continent Ecosystem (CCE) from 1962 to 2014. Specific objectives of this study were to determine how selected biophysical variables (slope, aspect, elevation, and latitude), regional climate variables (temperature, precipitation, and drought) and global climate oscillations (ENSO, PDO, NAO, AO, and PNA) relate to bark beetle infestations in the Crown of the Continent Ecosystem from 1962 through 2014, as measured by aerial surveyrecorded tree mortality. We sought to contextualize the results of the statistical models with historical data to further understand the relationship between increases and decreases of tree mortality by comparing these trends to geopotential height and sea-surface temperatures that may influence CCE climate. Our work revealed first, that while the aerial survey data has important limitations, overall it is a useful dataset for analyzing historical spatio-temporal patterns of insect infestations. Second, there appears to be a link between local biophysical factors, such as latitude, elevation, and winter precipitation (as opposed to global climate factors) and tree mortality within the CCE. Local climate analysis revealed the importance of winter precipitation to be the biggest influence of MPB decrease or increase along with lower geopotential heights during a decline in MPB spread over the CCE. Finally, a combination of a negative PDO and El Niño was important in forecasting a decline in MPB spread, as shown by damage, during a given year. This is the first study to use aerial survey data in a geospatial analysis incorporating biophysical variables for the US portion of the Crown of the Continent Ecosystem. Additionally, this study is unique to explore the potential relationship between global teleconnections and regional climate in the CCE area, and the spatio-temporal extent of mountain pine beetle infestations.Urban trees have become a key resource in building sustainable communities. Knowledge of preferences for trees, costs and benefits of trees, and how trees are managed by private and public landowners, is indispensable in making targeted planning that would fulfill the economic, social and environmental objectives of the urban communities This dissertation explores on these topics in three papers. Chapter 1 summarizes the three topics and briefly describes the motivation, methods applied and main conclusions in each study. Chapter 2 presents a meta-analysis that combines the non-market value estimates for tree cover from previous hedonic studies from various locations in the U.S. along with auxiliary data to investigate the relationship between the level of tree cover and value estimates. The study found that on average 35% and 40% tree cover respectively at property-level and county-level optimize the benefits to the property owners. Chapter 3 analyses how home owners’ preferences for trees vary across the landscape in multiple cities in the U.S., and the results are systematically summarized via an internal-meta analysis. The study confirms the preference heterogeneity across the landscape and found that the relative abundance of tree cover in study locations, regional differences and household characteristics affect the preferences. Chapter 4 develops a dynamic optimization model to study how private and public land ownership in local municipalities affects the cooperative management of urban trees to control the Emerald ash borer (EAB) infestation in Twin Cities, Minnesota. The results suggest that when an infested municipality has more public lands and when the transfer payments are efficiently used to implement greater control, the municipalities are more likely to commit to bargaining agreements that last for longer period of time. Results from the three studies provide insights to forward-looking communities to adapt their tree planting and protection efforts to lessen climate-induced impacts. The last chapter concludes the three studies and discusses the insights for future research.

Yes / This paper considers recent field evidence to analyse what online public services citizens need, explores potential citizen subsidy of these specific services and investigates where resources should be invested in terms of media accessibility. We explore these from a citizen-centric affordability perspective within three ‘exemplar’ developing countries in sub-Saharan Africa. The World Bank and United Nations in particular promote initiatives under the ‘Information and Communication Technologies for Development’ (ICT4D) to stress the relevance of e-Government as a way to ensure development and reduce poverty. We adopt a ‘Contingency Value’ method to conceptually outline reported citizens willingness to pay for digital public services. Hence, our focus is mainly upon an empirical investigation through extensive fieldwork in the context of sub-Sahara Africa. A substantive survey was conducted in the respective cities of Addis Ababa (Ethiopia), Lagos (Nigeria) and Johannesburg (South Africa). The sample of citizens was drawn from each respective Chamber of Commerce database for Ethiopia and South Africa, and for Nigeria a purchased database of businesses, based on stratified random sampling. These were randomly identified from both sectors ensuring all locations were covered with a total sample size of 1,297 respondents. It was found, in particular, that citizens were willing to pay to be able to access digital public services and that amounts of fees they were willing to pay varied depending on what services they wish to access and what devices they use (PCs or mobile phones). / European Investment Bank under the EIB-Universities Research Action Programme

Arithmetic Asian options are difficult to price and hedge, since, at the present, no closed-form analytical solution exists to price them. This difficulty, moreover, has led to the development of various methods and models used to price these instruments. The purpose of this thesis is two-fold. First, we present an overview of the literature. Secondly, we develop a pseudo-analytical method proposed by Geman and Yor and present an accurate and relatively quick algorithm which can be used to price European-style arithmetic Asian options and their hedge parameters. / Master of Science

In this work, we propose ModelPred, a framework that helps to understand the impact of changes in training data on a trained model. This is critical for building trust in various stages of a machine learning pipeline: from cleaning poor-quality samples and tracking important ones to be collected during data preparation, to calibrating uncertainty of model prediction, to interpreting why certain behaviors of a model emerge during deployment. Specifically, ModelPred learns a parameterized function that takes a dataset S as the input and predicts the model obtained by training on S. Our work differs from the recent work of Datamodels as we aim for predicting the trained model parameters directly instead of the trained model behaviors. We demonstrate that a neural network-based set function class is capable of learning the complex relationships between the training data and model parameters. We introduce novel global and local regularization techniques to prevent overfitting and we rigorously characterize the expressive power of neural networks (NN) in approximating the end-to-end training process. Through extensive empirical investigations, we show that ModelPred enables a variety of applications that boost the interpretability and accountability of machine learning (ML), such as data valuation, data selection, memorization quantification, and model calibration. / Amazon-Virginia Tech Initiative in Efficient and Robust Machine Learning / Master of Science / Also published as Zeng, Y., Wang, J. T., Chen, S., Just, H. A., Jin, R., & Jia, R. (2023, February). ModelPred: A Framework for Predicting Trained Model from Training Data. In 2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) (pp. 432-449). IEEE. https://doi.org/10.1109/SaTML54575.2023.00037 / With the prevalence of large and complicated Artificial Intelligence (AI) models, it is important to build trust in the various stages of a machine learning model pipeline, from cleaning poor-quality samples and tracking important ones to be collected during the training data preparation, to calibrating uncertainty of model prediction during the inference stage, to interpreting why certain behaviors of a model emerge during deployment. In this work, we propose ModelPred, a framework that helps to understand the impact of changes in training data on a trained model. To achieve this, ModelPred learns a parameterized function that takes a dataset S as the input and predicts the model obtained by training on S, thus learning the impact from data on the model efficiently. Our work differs from the recent work of Datamodels [28] as we aim for predicting the trained model parameters directly instead of the trained model behaviors. We demonstrate that a neural network-based set function class is capable of learning the complex relationships between the training data and model parameters. We introduce novel global and local regularization techniques to enhance the generalizability and prevent overfitting. We also rigorously characterize the expressive power of neural networks (NN) in approximating the end-to-end training process. Through extensive empirical investigations, we show that ModelPred enables a variety of applications that boost the interpretability and accountability of machine learning (ML), such as data valuation, data selection, memorization quantification, and model calibration. This greatly enhances the trustworthy of machine learning models.

The discounted cash flow model and relative valuation models are ever-increasingly prevalent in today’s investment-heavy environment. In other words, theoretically inferior models are used in practice. It is this paradox that has lead us to compare the discounted cash flow model (DCFM), discounted dividend model (DDM), residual income-based model (RIVM) and the abnormal earnings growth model (AEGM) and their relative accuracy to observed stockprices. Adding to previous research, we investigate their performance in relation to the OMX30 index. What is more, we test how the performance of each model is affected by an extension of the forecast horizon. The study finds that AEGM outperforms the other models, both before and after extending the horizon. Our analysis was conducted by looking at accuracy, spread and the inherent speculative nature of each model. Taking all this into account, RIVM outperforms the other models. In this sense, one can question the rationale behind investor’s decision to primarily use the discounted cash flow model in equity valuation.

published_or_final_version / Urban Planning / Master / Master of Science in Urban Planning