Browsing by Subject "cluster analysis"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Bacterial total maximum daily load (TMDL): development and evaluation of a new classification scheme for impaired waterbodies of Texas(Texas A&M University, 2005-02-17) Paul, SabuUnder the Clean Water Act (CWA) program the Texas Commission on Environmental Quality (TCEQ) listed 110 stream segments with pathogenic bacteria impairment in 2000. The current study was conducted to characterize the watersheds associated with the impaired waterbodies. The main characteristics considered for the classification of waterbodies were designated use of the waterbody, land use distribution, density of stream network, average distance of a land of a particular use to the closest stream, household population, density of on-site sewage facilities (OSSF), bacterial loading due to the presence of different types of farm animals and wildlife, and average climatic conditions. The availability of observed in-stream fecal coliform bacteria concentration data was evaluated to obtain subgroups of data-rich and data-poor watersheds within a group. The climatic data and observed in-stream fecal coliform bacteria concentrations were analyzed to find out seasonal variability of the water quality. The watershed characteristics were analyzed using the multivariate statistical analysis techniques such as factor analysis/principal component analysis, cluster analysis, and discriminant analysis. Six groups of watersheds were formed as result of the statistical analysis. The main factors that differentiate the clusters were found to be bacterial contribution from farm animals and wildlife, density of OSSF, density of households connected to public sewers, and the land use distribution. Two watersheds were selected each from two groups of watersheds. Hydrological Simulation Program-FORTRAN (HSPF) model was calibrated for one watershed within each group and tested for the other watershed in the same group to study the similarity in the parameter sets due to the similarity in watershed characteristics. The study showed that the watersheds within a given cluster formed during the multivariate statistical analysis showed similar watershed characteristics and yielded similar model results for similar model input parameters. The effect of parameter uncertainty on the in-stream bacterial concentration predictions by HSPF was evaluated for the watershed of Salado Creek, in Bexar County. The parameters that control the HSPF model hydrology contributed the most variance in the in-stream fecal coliform bacterial concentrations corresponding to a simulation period between 1 January 1995 and 31 December 2000.Item Forming Peer Advisory Groups in Agriculture: An Alternative Application of Cluster Analysis(2012-07-16) Doerr, Kayla MarieA "peer advisory group" essentially melds a business advisory board with a peer group. Peer advisory groups consist of business managers who meet together for the purpose of mutual self-improvement and learning through the sharing of experiences. The entire peer advisory group concept encompasses many variations and this research focuses on groups consisting of farm managers. Unfortunately, some farm managers who wish to participate have expressed frustration with group formation: they find it difficult to identify suitable individuals to participate in a peer advisory group with. Peer advisory groups can take many forms, and experts have suggested an individual should specifically seek out people interested in the same type of group. For example, an individual who wants to strictly focus discussion on production issues should seek out other individuals who also seek to focus on production discussions. Some individuals have suggested that some type of "clearinghouse" organizations could be beneficial in assisting individuals with the peer advisory group formation problem. Such an organization would likely need to adapt some sort of method for identifying individuals who have interest in a similar type of group. Although this could be approached from several different angles, one possible approach involves the practice of cluster analysis?a wide set of procedures intended to break down a set of objects into "clusters" of individuals with similar attributes. Cluster analysis comes with several attractive benefits; however, literature includes countless variations in the methods and criticisms of certain aspects of the methodology. This thesis focuses on using cluster analysis to assist with peer advisory group formation. More specifically, this thesis seeks to answer the following question: how could a clearinghouse organization apply cluster analysis methods to a pool of candidates to effectively create peer advisory groups congruent to the individuals' needs and wants? An approach was proposed which differs slightly from traditional cluster analysis methods, and this was applied to a hypothetical pool of candidates, along with several control methods. The proposed approach was found to most effectively create peer advisory groups which fulfilled the desires of the individuals.