http://ir.bdu.edu.et/handle/123456789/1858 2026-08-27T20:12:18Z http://ir.bdu.edu.et/handle/123456789/17064 Comprehensive Statistical Analytics for Sorghum Genotype Selection Under Irrigated and Non-Irrigated Conditions Mulugeta, Tesfa Sorghum is a widely cultivated crop in arid and semi-arid regions, particularly in sub-Saharan African countries. Several factors have been influencing sorghum production, with genotype and genotype-by-environment interaction being the primary sources of variability in yield and yield-related traits. This dissertation aimed to identify the best statistical approach for sorghum genotype selection with high performance in grain yield and yield-related traits under non-irrigated and irrigated conditions. Specifically, the sorghum genotype selection with superior performance in grain yield under both conditions is done by treating the genotypes as fixed or random effects using parametric, nonparametric statistical models and unsupervised machine learning to perform the genotype selection with superior performance in major traits. The experiment used water availability as a treatment, and each replication within the treatment levels used a lattice square design for data collection. A design consisted of 14×14 square experimental units containing 196 genotypes, where each row of the square represented a block receiving 14 genotypes, from which phenotypic traits were measured for the analysis. Findings revealed that the sorghum genotypes with superior performance in grain yield employed different parametric models. A result of genotype selection using mean performance in analysis of variance, with an assumption of genotypes treated as fixed effects, which is less important compared to other parametric models considering the genotypes as random effects. The genotype selection with superior performance in grain yield using mixed-effects models was a more reliable selection compared with the classical model. Findings of genotype selection with superior performance in major traits using a distributional assumption-free statistical model were advisable to scan the genotypes performing best in various phenotypic traits. Classification of phenotypic traits and sorghum genotypes was implemented to investigate high-performing genotypes in major phenotypic traits. The outcomes of this study can help academics and researchers in plant breeding for sorghum genotype selection in phenotype traits using advanced models. This can serve as an alternative statistical analysis approach for plant breeders. 2026-06-01T00:00:00Z http://ir.bdu.edu.et/handle/123456789/16991 A Probabilistic and Data-Driven Approach to Childhood Under nutrition Transitions Getnet, Bogale Child undernutrition remains a persistent public health challenge in Ethiopia, contributing to significant developmental, health, and economic burdens. Despite national commitments to reduce its prevalence, rates of stunting, underweight, and wasting remain high, particularly among children in rural and low-income households. Previous studies have often examined undernutrition in isolation, without accounting for its concurrent forms or the complex pathways through which it evolves over time. This study aimed to analyze the dynamic nature of childhood undernutrition in Ethiopia, focusing on the transitions between nutritional states and the combined effects of social, economic, environmental, and maternal factors. The research also investigates which factors predict long-term nutritional outcomes and how interventions can be better targeted to high-risk populations. Methodologically, the research applied advanced statistical and computational tools to longitudinal panel data, including multistate Markov chain models for state transitions, dynamic Bayesian networks for causal exploration, machine learning, and deep learning techniques for long-term trend prediction. A sample of over 1,999 children, aged 1 to 15, was followed across five survey rounds (2002–2016), with data collected from five diverse regions in Ethiopia-capturing a broad representation of geographic and socioeconomic contexts. The findings revealed that undernutrition is frequently experienced in multiple forms simultaneously and is influenced by a constellation of interrelated factors-particularly maternal education, household wealth, and access to health services and sanitation. Results also showed that many children transitioned between undernutrition states over time, with some remaining chronically affected. Predictive models accurately identified high-risk groups and highlighted early adolescence as a critical window for intervention. These results offer valuable insights for Ethiopian health planners and policymakers, reinforcing the need for integrated, long-term, and equity-focused strategies. By uncovering dynamic patterns and causal pathways, this study contributes to both academic understanding and practical decision-making for child nutrition programs in Ethiopia and similar contexts 2026-06-01T00:00:00Z http://ir.bdu.edu.et/handle/123456789/16989 Causal Inference for Count Treatments on Ordinal Outcomes Ashagrie, Sharew In an observational study, the propensity score is widely used to adjust for confounding bias when the treatment is binary and, to some extent, for categorical and continuous treatments. This study introduced the application of count treatments to causal inference on ordinal outcomes. We compared different methods, including maximum likelihood for the family of count models, generalized boosted model (GBM), and covariate balancing propensity score (CBPS), to estimate generalized propensity score (GPS). In the comparison, we used effective sample size in combination with confounders balancing power and efficiency of treatment effect estimation. In the treatment effect estimation, confounders’ bias was controlled using inverse probability of treatment weighting (IPTW). Two outcome models, including marginal structural modeling and considering GPS as a covariate, were compared in terms of the efficiency of treatment effect estimation. To reduce the influence of extreme IPTW, we trimmed the GPS at the 1st and 99th percentiles. We also applied path-specific effects estimation under sequential mediators while estimating the count treatments effect on the ordinal outcomes. In this case, a composite weighting method was introduced to estimate path-specific effects under sequential mediators to weight each component of the structural equation modeling of mediators and outcome. The first mediator model was weighted by IPTW generated from the treatment model; the second mediator model was weighted by the composite IPTW generated from the treatment and the first mediator model; and the outcome model was weighted by the composite IPTW generated from the treatment and mediators. The proposed weighting method was compared with simulation-based G-computation in terms of treatment effect estimation. For real-life data, we used the number of antenatal care visits (ANC) as a count treatment and age-specific childhood vaccination as an ordinal outcome. We also used a simulation study to compare the proposed models in the study. The result shows that GBM-based weighting is better than other methods to maintain a high effective sample size and efficient count treatment effect estimation while having equivalent performance with other methods in terms of covariate balancing. Marginal structural modeling produced better treatment effect estimation than using GPS as a covariate. The proposed weighting method and G-computation based on simulation have comparable performance in terms of treatment effect estimation, despite the fact that the proposed weighting method is computationally efficient. The number of ANC visits has a positive effect on the likelihood of timely childhood vaccination. In viii the meantime, the effect was mediated by institutional delivery and newborn postnatal care, though the direct effect of ANC visits is higher than the indirect effects. 2026-06-01T00:00:00Z http://ir.bdu.edu.et/handle/123456789/15907 Determinants of Caesarean Section Delivery among Woman in Ethiopian Administrative Zones: An Application of Geo Additive model Yimam Ebrahim Background: Caesarian section (CS) delivers the pregnancy outcome by making an incision through the maternal abdomen and uterus. Although there is a wide variation of CS in Sub Saharan Africa, currently Ethiopia has a considerably lower CS delivery rate. The main objective of this study is examine determinants of CS among women in Ethiopia's administrative zone. Method: The recent 2019 Ethiopia Mini Demographic and Health Survey (EMDHS) dataset with 5,527 weighted samples of women was used. Data management was done using STATA version 17 software. To investigate the associated determinants of CS, we used Geoadditive model. Results: From this study prevalence of CS in Ethiopia was 5.44%. Awi, East Gojjam, South Wollo, North Shewa zones in Amhara region; all Addis Ababa zones; East Shewa in Oromia region Kembata Tambaro and Sidama zones from SNNP region were hotspot area for CS. Women with primary, secondary and higher educational level had higher odds of C- section with Posterior Odds Ratio (POR=1.5278, 95%CI=1.1990-1.9484), (POR=2.1336, 95%CI=1.3971-3.2583) and (POR=4.0382, 95%CI=2.2175-7.3537), respectively as compared to those in no education category. Having ANC visits of 1-3, 4th and above, were associated with higher odds of CS as (POR= 2.6783, (1.5001-4.8521)) and (POR=2.7042, (1.5479-4.9704)) respectively with 95%CI compared to their reference no ANC visit. Conclusion: The generalized geo-additive effects model enables simultaneous modeling of spatial correlation, heterogeneity and possible nonlinear effects of covariates. Predictors of CS delivery in Ethiopia include mother's educational level, child twin, religion, place of residence, ANC visit, pregnancy counseling, place of delivery, total number of children born, birth order, mother's current age, preceding birth interval, and mother's age at first birth. The study also revealed signifcant spatial variations on CS delivery among administrative zones. It also depicts that the presence of spatial structured effect had negative or positive effect for CS delivery in Ethiopia zones. Priority interventions should focus on rural women, promoting women's education, and encouraging pregnancy counseling to address this issue 2024-07-01T00:00:00Z