Matching Items (460)
Filtering by
- All Subjects: artificial intelligence
- Creators: Arizona State University
- Creators: Department of Economics
Description
Phosphorus (P), an essential element for life, is becoming increasingly scarce, and its global management presents a serious challenge. As urban environments dominate the landscape, we need to elucidate how P cycles in urban ecosystems to better understand how cities contribute to — and provide opportunities to solve — problems of P management. The goal of my research was to increase our understanding of urban P cycling in the context of urban resource management through analysis of existing ecological and socio-economic data supplemented with expert interviews in order to facilitate a transition to sustainable P management. Study objectives were to: I) Quantify and map P stocks and flows in the Phoenix metropolitan area and analyze the drivers of spatial distribution and dynamics of P flows; II) examine changes in P-flow dynamics at the urban agricultural interface (UAI), and the drivers of those changes, between 1978 and 2008; III) compare the UAI's average annual P budget to the global agricultural P budget; and IV) explore opportunities for more sustainable P management in Phoenix. Results showed that Phoenix is a sink for P, and that agriculture played a primary role in the dynamics of P cycling. Internal P dynamics at the UAI shifted over the 30-year study period, with alfalfa replacing cotton as the main locus of agricultural P cycling. Results also suggest that the extent of P recycling in Phoenix is proportionally larger than comparable estimates available at the global scale due to the biophysical characteristics of the region and the proximity of various land uses. Uncertainty remains about the effectiveness of current recycling strategies and about best management strategies for the future because we do not have sufficient data to use as basis for evaluation and decision-making. By working in collaboration with practitioners, researchers can overcome some of these data limitations to develop a deeper understanding of the complexities of P dynamics and the range of options available to sustainably manage P. There is also a need to better connect P management with that of other resources, notably water and other nutrients, in order to sustainably manage cities.
ContributorsMetson, Genevieve (Author) / Childers, Daniel (Thesis advisor) / Aggarwal, Rimjhim (Thesis advisor) / Redman, Charles (Committee member) / Arizona State University (Publisher)
Created2011
Description
This thesis pursues a method to deregulate the electric distribution system and provide support to distributed renewable generation. A locational marginal price is used to determine prices across a distribution network in real-time. The real-time pricing may provide benefits such as a reduced electricity bill, decreased peak demand, and lower emissions. This distribution locational marginal price (D-LMP) determines the cost of electricity at each node in the electrical network. The D-LMP is comprised of the cost of energy, cost of losses, and a renewable energy premium. The renewable premium is an adjustable function to compensate `green' distributed generation. A D-LMP is derived and formulated from the PJM model, as well as several alternative formulations. The logistics and infrastructure an implementation is briefly discussed. This study also takes advantage of the D-LMP real-time pricing to implement distributed storage technology. A storage schedule optimization is developed using linear programming. Day-ahead LMPs and historical load data are used to determine a predictive optimization. A test bed is created to represent a practical electric distribution system. Historical load, solar, and LMP data are used in the test bed to create a realistic environment. A power flow and tabulation of the D-LMPs was conducted for twelve test cases. The test cases included various penetrations of solar photovoltaics (PV), system networking, and the inclusion of storage technology. Tables of the D-LMPs and network voltages are presented in this work. The final costs are summed and the basic economics are examined. The use of a D-LMP can lower costs across a system when advanced technologies are used. Storage improves system costs, decreases losses, improves system load factor, and bolsters voltage. Solar energy provides many of these same attributes at lower penetrations, but high penetrations have a detrimental effect on the system. System networking also increases these positive effects. The D-LMP has a positive impact on residential customer cost, while greatly increasing the costs for the industrial sector. The D-LMP appears to have many positive impacts on the distribution system but proper cost allocation needs further development.
ContributorsKiefer, Brian Daniel (Author) / Heydt, Gerald T (Thesis advisor) / Shunk, Dan (Committee member) / Hedman, Kory (Committee member) / Arizona State University (Publisher)
Created2011
Description
This dissertation consists of two essays. The first measures the degree to which schooling accounts for differences in industry value added per worker. Using a sample of 107 economies and seven industries, the paper considers the patterns in the education levels of various industries and their relative value added per worker. Agriculture has notably less schooling and is less productive than other sectors, while a group of services including financial services, education and health care has higher rates of schooling and higher value added per worker. The essay finds that in the case of these specific industries education is important in explaining sector differences, and the role of education all other industries are less defined. The second essay provides theory to investigate the relationship between agriculture and schooling. During structural transformation, workers shift from the agriculture sector with relatively low schooling to other sectors which have more schooling. This essay explores to what extent changes in the costs of acquiring schooling drive structural transformation using a multi-sector growth model which includes a schooling choice. The model is disciplined using cross country data on sector of employment and schooling constructed from the IPUM International census collection. Counterfactual exercises are used to determine how much structural transformation is accounted for by changes in the cost of acquiring schooling. These changes account for small shares of structural transformation in all economies with a median near zero.
ContributorsSchreck, Paul (Author) / Herrendorf, Berthold (Committee member) / Lagakos, David (Committee member) / Schoellman, Todd (Committee member) / Arizona State University (Publisher)
Created2011
Description
The global demand and trade for fruits and vegetables is increasing at national and international levels. The fresh fruits and vegetables supply chain are highly vulnerable to contamination and can be easily spoiled due to their perishable nature. Due to increases in fresh fruit and vegetable trade shipment volume between countries, the fresh food supply chain area is the highly susceptible and frequently prone to food contamination. The inability of firms in the fresh food business to have a good supply chain visibility and tracking system is one of the prominent reasons for food safety failure. Therefore, in order to avoid food safety risk and to supply safe food to consumers, the firms need to have an efficient traceability system in their supply chain. Most of the research in the food supply chain area suggests the implementation of a highly efficient tracking system called RFID (Radio frequency identification) technology to firms in the food industry. The medium scale firms in the fresh food supply chain business are skeptical about implementing the RFID technology equipped traceability system due to its high cost of investment and low margins on fresh food sales. This research developed two methods to measure the probability of food safety risk in food supply chain. These methods use the information gain from RFID traceability systems as a tool to measure the amount of risk in the fresh food supply chain. The stochastic optimization model is applied in this study to determine the risk premium by investing in RFID technology over the electronic barcode traceability system. The results show that there is a reduction in buyer (Type II error) and seller risk (Type I error) for RFID technology employed traceability system compared to electronic barcode system. It is found from stochastic optimization results that there is a positive risk premium by investing in RFID traceability system over the current systems and suggests the implementation of RFID traceability system for complex medium scale fresh produce imports to reduce the food safety risks. This research encourages the food industries and government agencies to evaluate alternatives to update supply chain system with RFID technology.
ContributorsJanke, Deepak Kumar (Author) / Nganje, William (Thesis advisor) / Schmitz, Troy (Committee member) / Thor, Eric (Committee member) / Arizona State University (Publisher)
Created2011
Description
This dissertation transforms a set of system complexity reduction problems to feature selection problems. Three systems are considered: classification based on association rules, network structure learning, and time series classification. Furthermore, two variable importance measures are proposed to reduce the feature selection bias in tree models. Associative classifiers can achieve high accuracy, but the combination of many rules is difficult to interpret. Rule condition subset selection (RCSS) methods for associative classification are considered. RCSS aims to prune the rule conditions into a subset via feature selection. The subset then can be summarized into rule-based classifiers. Experiments show that classifiers after RCSS can substantially improve the classification interpretability without loss of accuracy. An ensemble feature selection method is proposed to learn Markov blankets for either discrete or continuous networks (without linear, Gaussian assumptions). The method is compared to a Bayesian local structure learning algorithm and to alternative feature selection methods in the causal structure learning problem. Feature selection is also used to enhance the interpretability of time series classification. Existing time series classification algorithms (such as nearest-neighbor with dynamic time warping measures) are accurate but difficult to interpret. This research leverages the time-ordering of the data to extract features, and generates an effective and efficient classifier referred to as a time series forest (TSF). The computational complexity of TSF is only linear in the length of time series, and interpretable features can be extracted. These features can be further reduced, and summarized for even better interpretability. Lastly, two variable importance measures are proposed to reduce the feature selection bias in tree-based ensemble models. It is well known that bias can occur when predictor attributes have different numbers of values. Two methods are proposed to solve the bias problem. One uses an out-of-bag sampling method called OOBForest, and the other, based on the new concept of a partial permutation test, is called a pForest. Experimental results show the existing methods are not always reliable for multi-valued predictors, while the proposed methods have advantages.
ContributorsDeng, Houtao (Author) / Runger, George C. (Thesis advisor) / Lohr, Sharon L (Committee member) / Pan, Rong (Committee member) / Zhang, Muhong (Committee member) / Arizona State University (Publisher)
Created2011
Description
Hydropower generation is one of the clean renewable energies which has received great attention in the power industry. Hydropower has been the leading source of renewable energy. It provides more than 86% of all electricity generated by renewable sources worldwide. Generally, the life span of a hydropower plant is considered as 30 to 50 years. Power plants over 30 years old usually conduct a feasibility study of rehabilitation on their entire facilities including infrastructure. By age 35, the forced outage rate increases by 10 percentage points compared to the previous year. Much longer outages occur in power plants older than 20 years. Consequently, the forced outage rate increases exponentially due to these longer outages. Although these long forced outages are not frequent, their impact is immense. If reasonable timing of rehabilitation is missed, an abrupt long-term outage could occur and additional unnecessary repairs and inefficiencies would follow. On the contrary, too early replacement might cause the waste of revenue. The hydropower plants of Korea Water Resources Corporation (hereafter K-water) are utilized for this study. Twenty-four K-water generators comprise the population for quantifying the reliability of each equipment. A facility in a hydropower plant is a repairable system because most failures can be fixed without replacing the entire facility. The fault data of each power plant are collected, within which only forced outage faults are considered as raw data for reliability analyses. The mean cumulative repair functions (MCF) of each facility are determined with the failure data tables, using Nelson's graph method. The power law model, a popular model for a repairable system, can also be obtained to represent representative equipment and system availability. The criterion-based analysis of HydroAmp is used to provide more accurate reliability of each power plant. Two case studies are presented to enhance the understanding of the availability of each power plant and represent economic evaluations for modernization. Also, equipment in a hydropower plant is categorized into two groups based on their reliability for determining modernization timing and their suitable replacement periods are obtained using simulation.
ContributorsKwon, Ogeuk (Author) / Holbert, Keith E. (Thesis advisor) / Heydt, Gerald T (Committee member) / Pan, Rong (Committee member) / Arizona State University (Publisher)
Created2011
Description
Haiti has witnessed high deforestation rates in recent decades, caused largely by the fuel needs of a growing population. The resulting soil loss is estimated to have contributed towards a decline in agricultural productivity of 0.5% -1.2% per year since 1997. Recent studies show the potential of biochar use through pyrolysis technology to increase crop yields and improve soil health. However, the appropriateness of this technology in the context of Haiti remains unexplored. The three objectives of this research were to identify agricultural- and fuel-use-related needs and gaps in rural Haitian communities; determine the appropriateness of biochar pyrolyzer technology, used to convert agricultural biomass into a carbon-rich charcoal; and develop an action-oriented plan for use by development organizations, communities, and governmental institutions to increase the likelihood of adoption. Data were collected using participatory rural appraisal techniques involving 30 individual interviews and three focus-group discussions in the villages of Cinquantin and La Boule in the La Coupe region of central Haiti. Topics discussed include agricultural practices and assets, fuel use and needs, technology use and adoption, and social management practices. The Sustainable Livelihoods framework was used to examine the assets of households and the livelihood strategies being employed. Individual and focus group interviews were analyzed to identify specific needs and gaps. E.M. Rogers' Diffusion of Innovations theory was used to develop potential strategies for the introduction of pyrolysis technology. Preliminary results indicate biochar pyrolysis has potential to address agricultural and fuel needs in rural Haiti. Probable early adopters of biochar technology include households that have adopted new agricultural techniques in the past, and those with livestock. Education about biochar, and a variety of pyrolysis technology options from which villagers may select, are important factors in successful adoption of biochar use. A grain mill as an example in one of the study villages provides a model of ownership and use of pyrolysis technology that may increase its likelihood of successful adoption. Additionally, women represent a group that may be well suited to control a new local biochar enterprise, potentially benefiting the community.
ContributorsDelaney, Michael Ryan (Author) / Aggarwal, Rimjhim (Thesis advisor) / Chhetri, Nalini (Committee member) / Henderson, Mark (Committee member) / Arizona State University (Publisher)
Created2011
Description
Real-world environments are characterized by non-stationary and continuously evolving data. Learning a classification model on this data would require a framework that is able to adapt itself to newer circumstances. Under such circumstances, transfer learning has come to be a dependable methodology for improving classification performance with reduced training costs and without the need for explicit relearning from scratch. In this thesis, a novel instance transfer technique that adapts a "Cost-sensitive" variation of AdaBoost is presented. The method capitalizes on the theoretical and functional properties of AdaBoost to selectively reuse outdated training instances obtained from a "source" domain to effectively classify unseen instances occurring in a different, but related "target" domain. The algorithm is evaluated on real-world classification problems namely accelerometer based 3D gesture recognition, smart home activity recognition and text categorization. The performance on these datasets is analyzed and evaluated against popular boosting-based instance transfer techniques. In addition, supporting empirical studies, that investigate some of the less explored bottlenecks of boosting based instance transfer methods, are presented, to understand the suitability and effectiveness of this form of knowledge transfer.
ContributorsVenkatesan, Ashok (Author) / Panchanathan, Sethuraman (Thesis advisor) / Li, Baoxin (Committee member) / Ye, Jieping (Committee member) / Arizona State University (Publisher)
Created2011
Description
The current study was a benefit cost analysis that examined mental and behavioral health and prescription drug service use data of 347 participants (212 youth and 135 caregivers) from a bereavement intervention, the Family Bereavement Program (FBP).The preliminary goals of the current study were to compare the FBP intervention and the Literature Control (LC) groups at the six year follow-up on: (a) number of participants using mental/behavioral health services and prescription drugs, (b) the frequency of use of mental/behavioral health services and prescription drugs, and (c) the costs of mental/behavioral health services and prescription drugs. The final, and primary goal, was to (d) calculate the benefits of the FBP by analyzing the monetary difference between the LC and FBP groups in terms of cost of services used and then by applying those benefits to the cost of the intervention. Data representing participating youths' and caregivers' mental health service use and prescription drug use at the sixth year post-intervention were collected, as were the costs of those services. Results indicated that fewer FBP participants used services and prescription drugs than the Literature Control (LC) participants, but FBP participants, particularly the youth, used some low intensity services more frequently whereas the LC youth used more intensive and costly services more frequently. Consequently, service costs were greater for participants in the LC group than for participants in the FBP group. The benefit cost ratio revealed that the FBP, as delivered, saved society between $.15 and $.27 in mental and behavioral health costs for every dollar spent on the intervention. Implications of these findings and directions for future research are discussed.
ContributorsPorter, Michèle M (Author) / Hanish, Laura D. (Thesis advisor) / Sandler, Irwin N. (Committee member) / Wolchik, Sharlene A (Committee member) / Johnson, William G. (Committee member) / Arizona State University (Publisher)
Created2011
Description
TaxiWorld is a Matlab simulation of a city with a fleet of taxis which operate within it, with the goal of transporting passengers to their destinations. The size of the city, as well as the number of available taxis and the frequency and general locations of fare appearances can all be set on a scenario-by-scenario basis. The taxis must attempt to service the fares as quickly as possible, by picking each one up and carrying it to its drop-off location. The TaxiWorld scenario is formally modeled using both Decentralized Partially-Observable Markov Decision Processes (Dec-POMDPs) and Multi-agent Markov Decision Processes (MMDPs). The purpose of developing formal models is to learn how to build and use formal Markov models, such as can be given to planners to solve for optimal policies in problem domains. However, finding optimal solutions for Dec-POMDPs is NEXP-Complete, so an empirical algorithm was also developed as an improvement to the method already in use on the simulator, and the methods were compared in identical scenarios to determine which is more effective. The empirical method is of course not optimal - rather, it attempts to simply account for some of the most important factors to achieve an acceptable level of effectiveness while still retaining a reasonable level of computational complexity for online solving.
ContributorsWhite, Christopher (Author) / Kambhampati, Subbarao (Thesis advisor) / Gupta, Sandeep (Committee member) / Varsamopoulos, Georgios (Committee member) / Arizona State University (Publisher)
Created2011