Probabilistic models have been successfully applied for a wide variety of problems, such as but not limited to information retrieval, computer vision, bio-informatics and speech processing. Probabilistic models allow us to encode our assumptions about the data in an elegant fashion and enable us to perform machine learning tasks such as classification and clustering in a principled manner. Probabilistic models for bio-acoustic data help in identifying interesting patterns in the data (for instance, the species-specific vocabulary), as well as species identification (classification) in recordings where the label is not available. The focus of this thesis is to develop efficient inference techniques for existing models, as well as develop probabilistic models tailored to bioacoustic data. First, we develop inference algorithms for the supervised latent Dirichlet allocation (LDA) model. We present collapsed variational Bayes, collapsed Gibbs sampling and maximum-a-posteriori (MAP) inference for parameter estimation and classification in supervised LDA. We provide an empirical evaluation of the trade-off between computational complexity and classification performance of the inference methods for supervised LDA, on audio classification (species identification in this context)as well as image classification and document classification tasks. Next, we present novel probabilistic models for bird sound recordings, that can capture temporal structure at different hierarchical levels, and model additional information such as the duration and frequency of vocalizations. We present a non-parametric density estimation technique for parameter estimation and show that the MAP classifier for our models can be interpreted as a weighted nearest neighbor classifier. We provide an experimental comparison between the proposed models and a support vector machine based approach, using bird sound recordings from the Cornell Macaulay library. / Graduation date: 2011 / Access restricted to the OSU Community at author's request from Dec. 30, 2010 - Dec. 30, 2011

No description available.

No description available.

No description available.

Arid and semi-arid ecosystems of the southwestern U.S. are experiencing major changes that have profound impacts for community structure and ecosystem function. First, these ecosystems are experiencing dramatic shifts in vegetation composition as a result of the invasion of non-native species. Second these ecosystems are predicted to undergo substantial shifts in climate regime, which include increases in the variability and frequency of extreme temperature and precipitation events. It is not well understood how these current and predicted changes will affect the physiological performance of different plant types in arid and semi-arid ecosystems. To address the effect of these changes, this dissertation focused on the photosynthetic response of a native and non-native grass species, and dominant shrub species to precipitation across contrasting soil surfaces in southeastern Arizona. The native and non-native grasses were exposed to wet and dry seasonal precipitation and responses to precipitation events ('pulses') were measured over the course of a summer growing season. To gain a mechanistic understanding of these patterns, the biochemical and diffusion limitations to photosynthetic function were measured over the course of a pulse period. Building on this foundation, natural stands of the non-native grass species were exposed to sequences of different sized pulse events. The physiological performance of a dominant shrub species, Larrea tridentata, was measured in order to determine the biochemical and diffusional constraints to photosynthetic function across seasons and contrasting soil surfaces. The results showed that leaf area development of these grass species affects water availability and time lags in photosynthetic response. Initial soil moisture conditions across contrasting soil surfaces influence the magnitude of photosynthetic response in grasses. Large photosynthetic responses of the non-native grass require large and consecutive precipitation pulses. Co-limitation of photosynthesis of Larrea tridentata by diffusion and biochemistry does not illustrate typical trends across seasons and soil surfaces. Overall results demonstrate the importance of determining the mechanisms responsible for observed leaf-level photosynthetic patterns across individual pulse events, seasons, and contrasting soil surfaces. This is especially important for predicting the magnitude of the response of plant communities in arid and semi-arid ecosystems to species invasions and changes in climate.

No description available.

No description available.

No description available.

No description available.

No description available.