7141 research outputs found
Sort by
Attention-Based Learning of Tabular Data
No full textDeep learning has tremendous success with text and image data using supervised learning. However, labeling real-world data is costly and often unavailable. Despite deep learning’s success, it struggles with tabular data classification and clustering. Tabular data is at the heart of many application domains where traditional machine learning thrives over deep learning. The overarching goal of this thesis is to bridge the gap of deep learning on tabular data. This is done by contributing to two objectives. First, we challenge the assumption of independent and identically distributed (i.i.d) data in machine learning. We use graph neural networks to capture relationships between samples. Additionally, we employ attention-based learning to prioritize specific features in tabular data. This approach relaxes the i.i.d assumption and enables us to leverage relationships between features and samples. Results reveal relaxing i.i.d assumptions beat traditional methods in six out of ten datasets. Second, we explore the feasibility of deep learning for clustering. To this end, we propose a novel deep clustering method that incorporates attention. Results from clustering accuracy tested on sixteen tabular datasets demonstrate the effectiveness of between-feature attention for deep clustering. Furthermore, our method outperforms existing deep clustering methods, bringing deep clustering closer to traditional methods. The results of this thesis show that relaxing i.i.d assumptions leads to improved representation learning of tabular data, as demonstrated by better classification and clustering performance. However, traditional machine learning remains competitive. We discuss the limitations of deep learning, our proposed method, and directions for future work to improve tabular data representation
Decoding the Impact of Nitrogen Forms on Plant Growth in Soilless Culture
No full textPlants can utilize two major nitrogen forms: ammonium (NH4) and nitrate (NO3). Plant’s response to N forms varies with genotypes, relative proportions of N forms (NH4:NO3), N rate, pH and growing medium composition. Growth and physiological responses of 16 genotypes (1 cereal, 2 cash crops, 4 vegetables, 4 fruits and 5 ornamentals) were investigated under multiple NH4:NO3 proportions, N rates, pH levels and soilless growing media. All these factors significantly influenced growth and physiological parameters of selected genotypes. Moreover, growth and productivity were significantly superior in exclusive NO3 (100%) nutrition than 100% NH4. Majority of genotypes had optimal performance at pH 5.5-5.8 and N-rate of 7.5 mM. Moreover, N forms were more influential than N rates. In addition, sensitiveness to NH4 varied with genotypes and growing medium. In solution culture, 10% NH4 significantly reduced growth and developed NH4 stress. Stunted growth, reduced biomass and leaf area, foliage chlorosis, necrosis, abscission, and modification of root morphology were the symptoms of NH4 stress. Even though, biomass was reduced, foliage symptoms and root morphology modification did not occur @ 20% NH4 in hybrid culture (peatlite and solution). Similarly, in peatlite, 50-75% NH4 did not substantially reduced growth. Our findings suggest that N forms and their relative proportions are genotype-specific and varies with growing medium. Moreover, optimal NH4:NO3 proportion would improve crop NUE and enhance precision in nitrogen management and sustainability of cropping system
Exploring Relationships Between Sown to Grow, Academic Growth, and Attendance
No full textThe purpose of this quantitative correlational study was to explore if and to what extent a correlation existed between the implementation of the Sown to Grow social and emotional learning curriculum, academic growth, and attendance. The study was grounded in the CASEL framework used by district leaders and curriculum designers to guide the development of social and emotional learning programs in schools. The research questions examined the correlation between implementation, academic growth, and attendance. The sample consisted of 117 K-12 schools from an urban school district in the South. Archival implementation, academic growth, and attendance data were provided by the district and analyzed using Jeffreys\u27s Amazing Statistics Program (JASP). Statistical testing included Spearman’s correlation to assess the relationship between each paired variable from the research questions. The results showed positive but weak correlations between implementation data, academic growth, and attendance. Data were analyzed using Kruskal-Wallis one-way ANOVA to assess the difference between variables in the research questions. The results showed that there was no difference in the levels of implementation, academic growth, and attendance. The findings of this study suggest that the level of implementation of Sown to Grow of schools in an urban school district is not significantly correlated to academic growth and attendance. The findings also suggest no difference in the levels of implementation for academic growth and attendance. While the findings related to academic growth and attendance were limited it did provide insight on the fact that other factors contribute to student outcomes. This foundational study paves the way for further investigation into the efficacy of school-based social and emotional programs and its relationship to student outcomes
Effects of Economic Advantage and Race/Ethnicity on High School Graduation Rates During COVID-19
No full textThe purpose of this quantitative correlational study was to determine the moderating effects of economic advantage and race/ethnicity on graduation rates during COVID-19. Data collected from the Tennessee Department of Education’s Graduation Cohort Rates was examined for a numerical relationship between the variable of race/ethnicity with graduation rates and the variable of economic advantage or disadvantage and graduation rates. While an a priori power analysis determined a minimum sample size of 107, the final sample was 429. The data analyzed for a hierarchical regression proved that racial demographics (Asian/Black/Hispanic compared to White) at the school level were statistically significantly related to high school graduation rates for the 2020-2021 school year. Additionally, a hierarchical regression proved statistically significant results between students from economically disadvantaged homes and those from economically advantaged homes and the reported graduation rates. Recommendations for educational leaders to focus on targeted support and inclusive environments emphasize the need for systemic changes to address these disparities. The findings contributed to a deeper understanding of the factors affecting high school graduation rates and underscored the importance of targeted interventions to support all students’ success
20230813_114816~2
No full texthttps://digitalscholarship.tnstate.edu/m-chamberlain-landscapes/1015/thumbnail.jp
20230114_150042~2
No full texthttps://digitalscholarship.tnstate.edu/m-chamberlain-landscapes/1013/thumbnail.jp
0902191616
No full texthttps://digitalscholarship.tnstate.edu/m-chamberlain-landscapes/1001/thumbnail.jp