Geography and Genetics of Ecological Speciation in Pea Aphids

During ecological speciation, divergent natural selection drives evolution of ecological specialization and genetic differentiation of populations on alternate environments. Populations diverging onto the same alternate environments may be geographically widespread, so that divergence may occur at an array of locations simultaneously. Spatial variation in the process of divergence may produce a pattern of differences in divergence among locations called the Geographic Mosaic of Divergence. Diverging populations may vary in their degree of genetic differentiation and ecological specialization among locations. My dissertation examines the pattern and evolutionary processes of divergence in pea aphids (Acyrthosiphon pisum) on alfalfa (Medicago sativa) and clover (Trifolium pretense). In Chapter One, I examined differences among North American aphid populations in genetic differentiation at nuclear, sequence-based markers and in ecological specialization, measured as aphid fecundity on each host plant. In the East, aphids showed high host-plant associated ecological specialization and high genetic differentiation. In the West, aphids from clover were genetically indistinguishable from aphids on alfalfa, and aphids from clover were less specialized. Thus, the pattern of divergence differed among locations, suggesting a Geographic Mosaic of Divergence. In Chapter Two, I examined genomic heterogeneity in divergence in aphids on alfalfa and clover across North America using amplified fragment length polymorphisms (AFLPs). The degree of genetic differentiation varied greatly among markers, suggesting that divergent natural selection drives aphid divergence in all geographic locations. Three of the same genetic markers were identified as evolving under divergent selection in the eastern and western regions, and additional divergent markers were identified in the East. In Chapter Three, I investigated population structure of aphids in North America, France, and Sweden using AFLPs. Aphids on the same host plant were genetically similar across many parts of their range, so the evolution of host plant specialization does not appear to have occurred independently in every location. While aphids on alfalfa and clover were genetically differentiated in most locations, aphids from alfalfa and clover were genetically similar in both western North America and Sweden. High gene flow from alfalfa onto clover may constrain divergence in these locations.

The potential of natural, photosynthetic pigments to improve the efficiency of dye-sensitized solar cells

Gemstone Team Grenergy

SEQUENTIAL DECISIONS AND PREDICTIONS IN NATURAL LANGUAGE PROCESSING

Natural language processing has achieved great success in a wide range of ap- plications, producing both commercial language services and open-source language tools. However, most methods take a static or batch approach, assuming that the model has all information it needs and makes a one-time prediction. In this disser- tation, we study dynamic problems where the input comes in a sequence instead of all at once, and the output must be produced while the input is arriving. In these problems, predictions are often made based only on partial information. We see this dynamic setting in many real-time, interactive applications. These problems usually involve a trade-off between the amount of input received (cost) and the quality of the output prediction (accuracy). Therefore, the evaluation considers both objectives (e.g., plotting a Pareto curve). Our goal is to develop a formal understanding of sequential prediction and decision-making problems in natural language processing and to propose efficient solutions. Toward this end, we present meta-algorithms that take an existent batch model and produce a dynamic model to handle sequential inputs and outputs. Webuild our framework upon theories of Markov Decision Process (MDP), which allows learning to trade off competing objectives in a principled way. The main machine learning techniques we use are from imitation learning and reinforcement learning, and we advance current techniques to tackle problems arising in our settings. We evaluate our algorithm on a variety of applications, including dependency parsing, machine translation, and question answering. We show that our approach achieves a better cost-accuracy trade-off than the batch approach and heuristic-based decision- making approaches. We first propose a general framework for cost-sensitive prediction, where dif- ferent parts of the input come at different costs. We formulate a decision-making process that selects pieces of the input sequentially, and the selection is adaptive to each instance. Our approach is evaluated on both standard classification tasks and a structured prediction task (dependency parsing). We show that it achieves similar prediction quality to methods that use all input, while inducing a much smaller cost. Next, we extend the framework to problems where the input is revealed incremen- tally in a fixed order. We study two applications: simultaneous machine translation and quiz bowl (incremental text classification). We discuss challenges in this set- ting and show that adding domain knowledge eases the decision-making problem. A central theme throughout the chapters is an MDP formulation of a challenging problem with sequential input/output and trade-off decisions, accompanied by a learning algorithm that solves the MDP.