Theses and Dissertations at Montana State University (MSU)
Permanent URI for this communityhttps://scholarworks.montana.edu/handle/1/732
Browse
2 results
Search Results
Item Multi- and many-objective factored evolutionary algorithms(Montana State University - Bozeman, College of Engineering, 2023) Peerlinck, Amy; Chairperson, Graduate Committee: John SheppardMulti-Objective Optimization (MOO) is the problem of optimizing two or more competing objectives, where problems dealing with more than three competing objectives are termed as Many-Objective (MaOO). Such problems occur naturally in the real world. For example, many engineering design problems have to deal with competing objectives, such as cost versus quality in product design. How do we handle these competing objectives? To answer this question, population-based meta-heuristic algorithms that find a set of Pareto optimal solutions have become a popular approach. However, with the increase in complexity of problems, a single population approach may not be the most efficient to solve MOO problems. For this reason, co-operative co-evolutionary algorithms (CCEA) are used, which split the population into subpopulations optimizing over subsets of variables that can now be optimized simultaneously. Factored Evolutionary Algorithms (FEA) extends CCEA by including overlap in the subpopulations. This dissertation extends FEA to MOO, thus creating the Multi-Objective FEA (MOFEA). We apply MOFEA to different problems in the MOO family with positive results; these problems include combinatorial and continuous benchmarks as well as problems in the real-world domain of Precision Agriculture. Furthermore, we investigate the influence of different grouping techniques on continuous large-scale, MOO, and MaOO problems to help guide research to use the appropriate techniques for specific problems. Based on these results, we find that some MaOO problems lead to large sets of non-dominated solutions. From this, an Objective Archive Management (OAM) strategy is presented that creates separate archives for each objective based on performance and diversity criteria. OAM successfully reduces large solution sets to a more manageable size to help end-users make more informed decisions. The presented research makes four main contributions to the field of Computer Science: the creation of a new Multi-Objective framework to create and use subpopulation in a co-operative manner including the ability to use overlapping populations, the analysis of different grouping strategies and their influence on continuous optimization in both large- scale and multi-objective optimization, the introduction of a post-optimization solution set reduction approach, and the inclusion of an environmental objective into a real-world Precision Agriculture application.Item Improving a precision agriculture on-farm experimentation workflow through machine learning(Montana State University - Bozeman, College of Engineering, 2019) Peerlinck, Amy; Chairperson, Graduate Committee: John SheppardReducing environmental impact while simultaneously improving net return of crops is one of the key goals of Precision Agriculture (PA). To this end, an on-farm experimentation workflow was created that focuses on reducing the applied nitrogen (N) rate through variable rate application (VRA). The first step in the process, after gathering initial data from the farmers, creates experimental randomly stratified N prescription maps. One of the main concerns that arises for farmers within these maps is the large jumps in N rate between consecutive cells. To this end we successfully develop and apply a Genetic Algorithm to minimize rate jumps while maintaining stratification across yield and protein bins. The ultimate goal of the on-farm experiments is to determine the final N rate to be applied. This is accomplished by optimizing a net return function based on yield and protein prediction. Currently, these predictions are often done with simple linear and non-linear regression models. Our work introduces six different machine learning (ML) models for improving this task: a single layer feed-forward neural network (FFNN), a stacked auto-encoder (SAE), three different AdaBoost ensembles, and a bagging ensemble. The AdaBoost and bagging methods each use a single layer FFNN as its weak model. Furthermore, a simple spatial analysis is performed to create spatial data sets, to better represent the inherent spatial nature of the field data. These methods are applied to four actual fields' yield and protein data. The spatial data is shown to improve accuracy for most yield models. It does not perform as well on the protein data, possibly due to the small size of these data sets, resulting in a sparse data set and potential overfitting of the models. When comparing the predictive models, the deep network performed better than the shallow network, and the ensemble methods outperformed both the SAE and a single FFNN. Out of the four different ensemble methods, bagging had the most consistent performance across the yield and protein data sets. Overall, spatial bagging using FFNNs as the weak learner has the best performance for both yield and protein prediction.