You are here
Search results
(1 - 2 of 2)
- Title
- Balancing convergence and diversity in evolutionary single, multi and many objectives
- Creator
- Seada, Haitham
- Date
- 2017
- Collection
- Electronic Theses & Dissertations
- Description
-
"Single objective optimization targets only one solution, that is usually the global optimum. On the other hand, the goal of multiobjective optimization is to represent the whole set of trade-off Pareto-optimal solutions to a problem. For over thirty years, researchers have been developing Evolutionary Multiobjective Optimization (EMO) algorithms for solving multiobjective optimization problems. Unfortunately, each of these algorithms were found to work well on a specific range of objective...
Show more"Single objective optimization targets only one solution, that is usually the global optimum. On the other hand, the goal of multiobjective optimization is to represent the whole set of trade-off Pareto-optimal solutions to a problem. For over thirty years, researchers have been developing Evolutionary Multiobjective Optimization (EMO) algorithms for solving multiobjective optimization problems. Unfortunately, each of these algorithms were found to work well on a specific range of objective dimensionality, i.e. number of objectives. Most researchers overlooked the idea of creating a cross-dimensional algorithm that can adapt its operation from one level of objective dimensionality to the other. One important aspect of creating such algorithm is achieving a careful balance between convergence and diversity. Researchers proposed several techniques aiming at dividing computational resources uniformly between these two goals. However, in many situations, only either of them is difficult to attain. Also for a new problem, it is difficult to tell beforehand if it will be challenging in terms of convergence, diversity or both. In this study, we propose several extensions to a state-of-the-art evolutionary many-objective optimization algorithm - NSGA-III. Our extensions collectively aim at (i) creating a unified optimization algorithm that dynamically adapts itself to single, multi- and many objectives, and (ii) enabling this algorithm to automatically focus on either convergence, diversity or both, according to the problem being considered. Our approach augments the already existing algorithm with a niching-based selection operator. It also utilizes the recently proposed Karush Kuhn Tucker Proximity Measure to identify ill-converged solutions, and finally, uses several combinations of point-to-point single objective local search procedures to remedy these solutions and enhance both convergence and diversity. Our extensions are shown to produce better results than state-of-the-art algorithms over a set of single, multi- and many-objective problems."--Pages ii-iii.
Show less
- Title
- Digital Evolution in Experimental Phylogenetics and Evolution Education
- Creator
- Kohn, Cory
- Date
- 2021
- Collection
- Electronic Theses & Dissertations
- Description
-
The creation and evaluation of known evolutionary histories and the implementation of student investigatory experiences on evolution are difficult endeavors that have only recently been feasible. The research presented in this dissertation is related in their shared use of digital evolution with Avidians as a model study system, both to conduct science research in experimental phylogenetics and to conduct education research in curricular intervention to aid student understanding.I first...
Show moreThe creation and evaluation of known evolutionary histories and the implementation of student investigatory experiences on evolution are difficult endeavors that have only recently been feasible. The research presented in this dissertation is related in their shared use of digital evolution with Avidians as a model study system, both to conduct science research in experimental phylogenetics and to conduct education research in curricular intervention to aid student understanding.I first present background discussions on the Avidian digital evolution study system—as implemented in Avida and Avida-ED—and its favorable use in experimental phylogenetics and biology education owing to its greater biological realism than computational simulations, and greater utility and generality than biological systems. Prior work on conducting experimental evolution for use in phylogenetics and work on developing undergraduate lab curricula using experimental evolution are also reviewed. I establish digital evolution as an effective method for phylogenetic inference validation by demonstrating that results from a known Avidian evolutionary history are concordant, under similar conditions, to established biological experimental phylogenetics work. I then further demonstrate the greater utility and generality of digital evolution over biological systems by experimentally testing how phylogenetic accuracy may be reduced by complex evolutionary processes operating singly or in combination, including absolute and relative degrees of evolutionary change between lineages (i.e., inferred branch lengths), recombination, and natural selection. These results include that directional selection aids phylogenetic inference, while stabilizing selection impedes it. By evaluating clade accuracy and clade resolvability across treatments, I evaluate measures of tree support and its presentation in the form of consensus topologies and I offer several general recommendations for systematists. Using a larger and more biologically realistic experimental design, I systematically examine a few of the complex processes that are hypothesized to affect phylogenetic accuracy—natural selection, recombination, and deviations from the model of evolution. By analyzing the substitutions that occurred and calculating selection coefficients for derived alleles throughout their evolutionary trajectories to fixation, I show that molecular evolution in these experiments is complex and proceeding largely as would be expected for biological populations. Using these data to construct empirical substitution models, I demonstrate that phylogenetic inference is incredibly robust to significant molecular evolution model deviations. I show that neutral evolution in the presence of always-occurring population processes, such as clonal or Hill-Robertson interference and lineage sorting, result in reduced clade support, and that selection and especially recombination, including their joint occurrence, restore this otherwise-reduced phylogenetic accuracy. Finally, this work demonstrates that inferred branch lengths are often quite inaccurate despite clade support being accurate. While phylogenetic inference methods performed relatively well in both theoretically facile and challenging molecular evolution scenarios, their accuracy in clade support might be a remarkable case of being right for misguided reasons, since branch length inference were largely inaccurate, and drastically different models of evolution made little difference. This work highlights the need for further research that evaluates phylogenetic methods under experimental conditions and suggests that digital evolution has a role here. Finally, I examine student understanding of the importance of biological variation in the context of a course featuring a digital evolution lab. I first describe the Avida-ED lab curriculum and its fulfillment of calls for reform in education. Then I describe the specific education context and other course features that aim to address student conceptualization of variation. I present a modified published assessment on transformational and variational understanding and findings regarding student understanding of variation within an evolution education progression. Finally, I offer suggestions on incorporating course material to engage student understanding of variation.
Show less