Estimating Regression Models Research Articles

Epidemiological survey of accidents with venomous animals treated at the Regional Hospital of Bom Jesus, Piauí, Northeast Brazil

Aim: Venomous animals are responsible for 26.8% of all poisonings occurring in people registered in Brazil, reflecting serious economic, medical, and social public health issues in tropical countries. The objective of this work was to characterize the cases notified of accidents involving venomous animals in humans in a Public Regional Hospital located in Piauí state, Northeast Brazil. Methods: We analyzed the notification forms registered at the National Notifiable Diseases Information System (SINAN), provided by the Hospital Regional Manoel de Souza from July 2009 to December 2017. The data were organized in spreadsheets and were exported to SPSS v. 24 for statistical analysis. A linear regression analysis was performed, and variables were considered significant when the estimated regression model presented P ≤ 0.05. Results: From 2009 to 2017, 833 cases of accidents caused by venomous animals treated were reported on SINAN. Accidents caused by scorpions were the most frequent, accounting for 69.7% (n = 269) of cases, followed by snakes at 17.6% (n = 68) and insects at 12.7% (n = 49). The people most affected by venomous animals were men (53.9%; n = 208) between the ages of 20 and 39 (42.5%; n = 164). The most frequent bite site was on the extremities, with the feet being the most affected region at 39.3% (n = 235), followed by the hands at 25.8% (n = 154) and the legs at 8% (n = 48). The study showed that despite the Regional Hospital covering 18 cities, the municipality of Bom Jesus had the highest incidence of accidents caused by venomous animals, mainly scorpions. Therefore, this investigation is a starting point for understanding the spatial distribution of these accidents. The data show that there are still flaws in the SINAN information flow. In the serotherapy variable, for example, 51.1% of cases were not completed. Conclusion: It is necessary to provide guidance and carry out educational work in order to improve official data records, so that they can be incorporated into the care routine.

Some new asymptotic results for series estimation under clustered dependence

This paper presents some asymptotic results for series estimation of a nonparametric regression model under clustered dependence. A mean square rate of convergence for the series regression estimator is established. Moreover, asymptotic pointwise normality is shown for the series estimator.

Cost Effectiveness of Fremanezumab in Episodic and Chronic Migraine Patients from a Japanese Healthcare Perspective.

Fremanezumab is an effective treatment for episodic (EM) and chronic migraine (CM) patients in Japan, but its cost effectiveness remains unknown. The objective of this study was to determine the cost effectiveness of fremanezumab compared with standard of care (SOC) in previously treated EM and CM patients from a Japanese healthcare perspective. Estimated regression models were implemented in a probabilistic Markov model to inform effectiveness and health-related quality-of-life data for fremanezumab and SOC. The model was further populated with data from the literature. The adjusted Japanese healthcare perspective included productivity losses. The main model outcomes were quality-adjusted life-years (QALYs), costs (2022 Japanese Yen [¥]), and incremental outcomes including the incremental cost-effectiveness ratio (ICER). Analyses were performed separately for the EM and CM patients and combined. Costs and effects were discounted at an annual rate of 2.0%. The mean QALYs over a 25-year time horizon for the EM and CM populations combined were 13.03 for SOC and 13.15 for fremanezumab. The associated costs were ¥27,550,292 for SOC and ¥28,371,048 for fremanezumab. QALYs were higher and costs lower for EM patients compared with CM patients for both fremanezumab and SOC. The deterministic ICERs of fremanezumab versus SOC were ¥6,334,861 for EM, ¥7,393,824 for CM, and ¥6,530,398 for EM and CM combined. Indirect costs and choice of mean migraine days model distribution had a substantial impact on the ICER. Using fremanezumab in a heterogeneous mixture of Japanese EM and CM patients resulted in a reduction of monthly migraine days and thus more QALYs compared with SOC. The cost effectiveness of fremanezumab versus SOC in EM and CM patients resulted in an ICER of ¥6,530,398, from an adjusted Japanese public healthcare perspective.

Clinical Efficacy Analysis of Wedge Resection of Pulmonary in Patients with Small Volume Invasive Lung Adenocarcinoma

With further understanding and research into non-small cell lung cancer with tumours ≤2 cm in maximum diameter, segmental lung resection is able to achieve the same long-term prognosis as lobectomy. However, there are few studies on the prognostic effect of wedge resection on small volume invasive lung adenocarcinoma with an invasion depth of 0.5 to 1.0 cm. Therefore, this study focuses on the clinical efficacy and prognosis of wedge resection in patients with small-volume invasive lung adenocarcinoma. A retrospective analysis of the medical records of 208 patients who underwent surgery in the Department of Thoracic Surgery of the Affiliated Provincial Hospital of Anhui Medical University from February 2016 to December 2017 was made, and the postoperative pathological results confirmed small volume invasive lung adenocarcinoma. According to their surgical methods, they were divided into lobectomy group (n=115), segmentectomy group (n=48) and wedge resection group (n=45). Kaplan-Meier survival curve estimation and Cox proportional risk regression model were used to explore the influence of different surgical methods on the prognosis of patients with small volume invasive lung adenocarcinoma. The wedge resection group had better perioperative outcomes compared with the segmentectomy group and lobectomy group, with statistically significant differences in intraoperative bleeding (P=0.036), postoperative drainage (P<0.001), operative time (P=0.018), postoperative time with tubes (P=0.001), and postoperative complication rate (P=0.006). There were no significant differences when comparing the three groups in terms of survival rate (lobectomy group vs segmentectomy group, P=0.303; lobectomy group vs wedge resection group, P=0.742; and segmentectomy group vs wedge resection group, P=0.278) and recurrence-free survival rate (lobectomy group vs segmentectomy group, P=0.495; lobectomy group vs wedge resection group, P=0.362; segmentectomy group vs wedge resection group, P=0.775). Univariate and multivariate survival analyses showed that consolidation tumor ratio (CTR) was the prognostic factor of overall survival and revurrence-free survival for patients with small-volume invasive lung adenocarcinoma (P<0.05). Wedge resection in patients with small volume invasive lung adenocarcinoma can achieve long-term outcomes similar to segmentectomy and lobectomy. When the CTR≤0.5, wedge resection is preferred in such patients.

Heavy Metal Concentration Estimation for Different Farmland Soils Based on Projection Pursuit and LightGBM with Hyperspectral Images.

Heavy metal pollution in farmland soil threatens soil environmental quality. It is an important task to quickly grasp the status of heavy metal pollution in farmland soil in a region. Hyperspectral remote sensing technology has been widely used in soil heavy metal concentration monitoring. How to improve the accuracy and reliability of its estimation model is a hot topic. This study analyzed 440 soil samples from Sihe Town and the surrounding agricultural areas in Yushu City, Jilin Province. Considering the differences between different types of soils, a local regression model of heavy metal concentrations (As and Cu) was established based on projection pursuit (PP) and light gradient boosting machine (LightGBM) algorithms. Based on the estimations, a spatial distribution map of soil heavy metals in the region was drawn. The findings of this study showed that considering the differences between different soils to construct a local regression estimation model of soil heavy metal concentration improved the estimation accuracy. Specifically, the relative percent difference (RPD) of As and Cu element estimations in black soil increased the most, by 0.30 and 0.26, respectively. The regional spatial distribution map of heavy metal concentration derived from local regression showed high spatial variability. The number of characteristic bands screened by the PP method accounted for 10-13% of the total spectral bands, effectively reducing the model complexity. Compared with the traditional machine model, the LightGBM model showed better estimation ability, and the highest determination coefficients (R2) of different soil validation sets reached 0.73 (As) and 0.75 (Cu), respectively. In this study, the constructed PP-LightGBM estimation model takes into account the differences in soil types, which effectively improves the accuracy and reliability of hyperspectral image estimation of soil heavy metal concentration and provides a reference for drawing large-scale spatial distributions of heavy metals from hyperspectral images and mastering soil environmental quality.

Neoadjuvant chemotherapy is associated with improved disease-free survival in pancreatic cancer patients undergoing pancreaticoduodenectomy with vascular resection.

Neoadjuvant chemotherapy (NAC) is becoming favored for all pancreatic adenocarcinoma (PDAC). Patients with seemingly resectable disease infrequently still display vascular involvement intraoperatively. Outcomes following NAC versus upfront surgery in patients undergoing pancreaticoduodenectomy (PD) with vascular resection are unknown. We performed a retrospective cohort study of PDAC patients who underwent PD with vascular resection between January 1, 2013, to December 31, 2020, within a single academic center. Clinicopathologic characteristics and disease-free survival (DFS) were compared between NAC versus upfront surgery cohorts using the Kaplan-Meier estimate and Cox proportional-hazards regression model. Eighty-one patients who underwent PD with vascular resection for PDAC were included. Forty-six patients (56%) received NAC. The NAC cohort more often had pathologic N0 status (47.8% vs. 8.6%, p < 0.001), had decreased vascular invasion (11% vs. 40%, p = 0.002), and completed chemotherapy (80% vs. 40%, p < 0.01). The NAC cohort demonstrated improved DFS (40.5 vs. 14.3 months, p = 0.007). In multivariable analysis, NAC remained independently associated with increased DFS (HR = 0.48, p = 0.02). NAC was associated with improved clinicopathologic outcomes and DFS in PD with vascular resection. These findings demonstrate the advantage of NAC in PDAC patients undergoing PD with vascular resection.

MSE superiority of the unrestricted Stein-rule estimator in a regression model with a possible structural break

This paper investigates the estimation of a linear regression model with a possible structural break at a known point. We analytically derive the exact formulae of the MSE for the restricted SR and PSR estimators, which shrinks the OLS estimator toward the restriction of no structural break. We compare the MSE performance of restricted/unrestricted SR, PSR, and least squared estimators. We analytically show that the unrestricted SR estimator can have a smaller MSE than the restricted SR estimator even when the restriction is correct. Further, our numerical results show that the unrestricted PSR estimator has the best MSE performance over a wide region of parameter space. These results indicate that the use of the unrestricted PSR estimator is recommended even when a structural break may not exist, although the unrestricted PSR estimator does not take the possibility of no structural break into consideration.

Abstract PO5-16-05: Integrating Circulating Tumor Cells (CTCs) and Cell-free DNA (cfDNA) Signatures for Monitoring Treatment Response in Stage III/IV Inflammatory Breast Cancer

Abstract Background: Inflammatory breast cancer (IBC) is a rare and aggressive form of breast cancer, comprising only between 1-5% of all breast cancer cases in the United States. Despite its low incidence, IBC accounts for 10% of breast cancer-related deaths and is often diagnosed with metastasis, resulting in shorter survival compared to non-IBC cases. The 5-year relative survival rate for all stages of IBC is 39%, dropping to 19% for patients with distant metastasis. Liquid biopsy-based biomarkers have emerged as potential analytes for predicting prognosis and monitoring response to therapies in IBC patients. We explored the prognostic role of liquid biopsy encompassing both CTCs and cfDNA in the blood of patients with stage III/IV IBC and their clinical outcomes. Methods: CTC and cfDNA analysis was performed on 90 blood samples from 47 patients with stage III and IV IBC. Samples were collected before (pre-treatment, n=43) initiation and after (post-treatment, n=47) completion of neoadjuvant chemotherapy. CTCs were enriched and enumerated using the CellSearch® platform (Menarini-Silicon Biosystems). CfDNA was extracted from plasma using MagMax™ Cell-Free Total Nucleic Acid Isolation Kit (Applied Biosystems™) and quantified using the Qubit™ dsDNA High Sensitivity kit (Invitrogen™). Differences in cfDNA concentrations were used to plot a receiver operating characteristic (ROC) curve which was then used to determine Area Under the Curve (AUC) and a cfDNA cut-off point. ROC cut-off was calculated based on Youden Index J. Survival analysis was performed by the Kaplan-Meier estimator and Cox regression models. P-values were measured using the Student t-test. Results: Prior to neoadjuvant treatment, CTC was detected in 51% of patient samples, and the median cfDNA concentration was 5.87 ng/mL (range 0.95-24.11 ng/mL). After completion of neoadjuvant treatment, CTC was detected in 25% of patient samples, and the median cfDNA concentration was 11.19 ng/mL (range 3.41-88.13 ng/mL). The cut-off points determined for pre-treatment and post-treatment cfDNA concentration was &amp;gt;4.2 ng/mL and &amp;gt;9.5 ng/mL respectively. At pre-treatment, CTC detection and cfDNA concentration &amp;gt;4.2 ng/mL were independently associated with disease relapse (P=0.007, 0.016, respectively). The combination of CTC detection and cfDNA concentration &amp;gt;4.2 ng/mL was significantly associated with disease relapse (P=0.0004), decreased relapse free survival (RFS, P=0.005) and overall survival (OS, P=0.007). At the post-neoadjuvant treatment time point, the combination of CTC detection and cfDNA concentration &amp;gt;9.5 ng/mL was significantly associated with relapse regardless of pCR status (P=0.02). Conclusion: The combined presence of CTCs and elevated cfDNA concentrations in patients with stage III/IV IBC after completion of neoadjuvant chemotherapy was associated with disease relapse regardless of pCR status. A combination of elevated cfDNA and CTC detection signatures after completion of neoadjuvant chemotherapy may be useful for risk stratification and adjuvant systemic therapy planning. Citation Format: Sridevi Addanki, Salyna Meas, Vanessa Sarli, Abdulkader Almosa, Wendy Woodward, Dhruvajyoti Roy, Anthony Lucci. Integrating Circulating Tumor Cells (CTCs) and Cell-free DNA (cfDNA) Signatures for Monitoring Treatment Response in Stage III/IV Inflammatory Breast Cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-16-05.

Logistic Regression Models for Sex Estimation Using Percutaneous Upper Arm and Forearm Anthropometric Dimensions in Adolescent School Children in Southwest Nigeria

Logistic Regression Models for Sex Estimation Using Percutaneous Upper Arm and Forearm Anthropometric Dimensions in Adolescent School Children in Southwest Nigeria

Pengelolaan Laba dalam Sektor Consumer Non-Cyclicals: Peran Aset Pajak Tangguhan, Skala Perusahaan, dan Leverage di Bursa Efek Indonesia

This research aims to empirically prove the influence of deferred tax assets, company size and leverage on earnings management. This research was conducted at a Consumer Non-Cyclical sector company. The research method used is quantitative. The type of data used is secondary in the form of annual financial reports published on the Indonesia Stock Exchange (BEI) for 2018-2022. Samples were collected using purposive sampling. The number of companies used as research samples was 28 companies with a research period of 5 (five) years, so that 140 observation data were obtained. Data processing uses the Eviews 12 Statistical Program by collecting related data then analyzing descriptive statistics of regression model estimates, model suitability tests, classical assumption tests, coefficient of determination (R 2) panel data regression analysis, F statistical tests and t statistical tests, Results The t statistical test for the variable deferred tax assets partially influences earnings management, but company size and leverage partially have no influence on earnings management. F test results: Deferred tax assets, company size and leverage simultaneously influence earnings management.

Bayes estimator for the inequality-constrained regression model*

This article studies the Bayesian estimation of the inequality-constrained regression model under a balanced loss function. The usual Bayes estimator is often involved in some complex integrals and has no explicit expression, which causes the corresponding properties be difficult to describe. Based on the idea of linear Bayes method suggested by Rao (1973), the present paper constructs a linear Bayes estimator for the regression parameters without specify their specific prior distributions. It is proved that the proposed linear Bayes estimator outperforms the corresponding least squares estimator, and numerical simulations show that when there is collinearity in design matrix, the linear Bayes estimator performs better than the usual Bayes estimator.

Traumatic brain injury incidence and mortality: a large population-based study.

Traumatic brain injury (TBI) is a severe condition that represents a major global public health concern. Provide a comprehensive epidemiological outlook encompassing TBI incidence, healthcare provision and mortality. Population-based study in Veneto (4.9 million inhabitants), Italy, from 2012 to 2021. Hospital discharge and mortality records were used to assess incidence and mortality. Kaplan-Meier survival estimator and Cox regression models were fitted to investigate determinants of mortality. Between 2012 and 2021, there were 37,487 incident TBI cases, corresponding to an age-standardized rate of 77.30/100,000 people (95% CI 76.52-78.09), higher among males, with an exponential growth after age 70. Leading causes were domestic (33.1%) and traffic accidents (17.7%), the first predominating among the elderly and children, while the latter in males 15-24 and older people. After rates stably declined between 2012 and 2019, the study captured a sharp decrease especially for traffic and occupational accidents in males, due to COVID-19 lockdown in 2020. Overall, 48.9% TBI patients were hospitalized in a specialized trauma center, with 2.6% requiring a transfer after accessing a spoke hospital. Over a 3.7 years median follow-up, 16,145 deaths were recorded, with higher mortality for those undergoing neurosurgical interventions, regardless of their access point. Risks of death increased with age, male gender, and comorbidities. TBI incidence is characterized by distinct patterns, affecting particularly older individuals and males. Minimal hospital transfers with comparable survival irrespective of access point suggests an effective patient management within the network. The study underscores the critical need for acute-phase support and prolonged care strategies for older TBI patients.

Overcoming Long Inference Time of Nearest Neighbors Analysis in Regression and Uncertainty Prediction

The intuitive approach of comparing like with like, forms the basis of the so-called nearest neighbor analysis, which is central to many machine learning algorithms. Nearest neighbor analysis is easy to interpret, analyze, and reason about. It is widely used in advanced techniques such as uncertainty estimation in regression models, as well as the renowned k-nearest neighbor-based algorithms. Nevertheless, its high inference time complexity, which is dataset size dependent even in the case of its faster approximated version, restricts its applications and can considerably inflate the application cost. In this paper, we address the problem of high inference time complexity. By using gradient-boosted regression trees as a predictor of the labels obtained from nearest neighbor analysis, we demonstrate a significant increase in inference speed, improving by several orders of magnitude. We validate the effectiveness of our approach on a real-world European Car Pricing Dataset with approximately 4.2×106\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$4.2 \ imes 10^6$$\\end{document} rows for both residual cost and price uncertainty prediction. Moreover, we assess our method’s performance on the most commonly used tabular benchmark datasets to demonstrate its scalability. The link is to github repository where the code is available: https://github.com/koutefra/uncertainty_experiments.

A Pretest Estimator for the Two-Way Error Component Model

For a panel data linear regression model with both individual and time effects, empirical studies select the two-way random-effects (TWRE) estimator if the Hausman test based on the contrast between the two-way fixed-effects (TWFE) estimator and the TWRE estimator is not rejected. Alternatively, they select the TWFE estimator in cases where this Hausman test rejects the null hypothesis. Not all the regressors may be correlated with these individual and time effects. The one-way Hausman-Taylor model has been generalized to the two-way error component model and allow some but not all regressors to be correlated with these individual and time effects. This paper proposes a pretest estimator for this two-way error component panel data regression model based on two Hausman tests. The first Hausman test is based upon the contrast between the TWFE and the TWRE estimators. The second Hausman test is based on the contrast between the two-way Hausman and Taylor (TWHT) estimator and the TWFE estimator. The Monte Carlo results show that this pretest estimator is always second best in MSE performance compared to the efficient estimator, whether the model is random-effects, fixed-effects or Hausman and Taylor. This paper generalizes the one-way pretest estimator to the two-way error component model.

Social Networks and Communication Impairment: Racial Differences Among Stroke Survivors

Purpose: Recent evidence highlights the importance of social networks—the composition of social relationships—and their characteristics in stroke recovery. This study explores the relationship between social networks and poststroke communication outcomes. Method: A cohort of 30 non-Hispanic Black ( n = 12) and non-Hispanic White ( n = 18) adult stroke survivors completed the Stroke Impact Scale (SIS) Communication subtest and the Stroke Social Network Scale (SSNS). The SSNS captures elements of social networks including satisfaction and contact with friends, relatives, children, and groups. The relationship between the SIS Communication subtest scores and SSNS scores was explored using ordinal Bayesian regression estimation models adjusting for sex, age, time postonset, marital status, and stroke type. Results: Average SIS Communication scores differed ( t = 2.07, p = .335) between White ( M = 81.75, SD = 22.90) and Black ( M = 95.71, SD = 4.95) participants, but SSNS Comprehensive scores were similar between the two groups ( t = −0.89, p = 1.00). SSNS children subtest scores were higher among the Black than the White ( t = −1.73, p = .0473) participants. Similarly, satisfaction subtest scores were comparatively higher among Black than White participants ( t = 1.85, p = .451). SSNS Satisfaction, Friends, and Groups subtest scores were positively associated with SIS Communication, indicating strong associations with lower impairment. SSNS Children subtest scores were negatively related to SIS, indicating higher impairment. Although most subtest associations were similar for both groups, the associations with SSNS Children and Friends subtests differed between racial groups. Conclusions: Levels of communication impairment were correlated with child, friendship, and group attachments among participants. However, strong social ties to children and friends had different associations among racial groups.

New ridge parameter estimators for the quasi-Poisson ridge regression model

The quasi-Poisson regression model is used for count data and is preferred over the Poisson regression model in the case of over-dispersed count data. The quasi-likelihood estimator is used to estimate the regression coefficients of the quasi-Poisson regression model. The quasi-likelihood estimator gives sub-optimal estimates if regressors are highly correlated—multicollinearity issue. Biased estimation methods are often used to overcome the multicollinearity issue in the regression model. In this study, we explore the ridge estimator for the quasi-Poisson regression model to mitigate the multicollinearity issue. Furthermore, we propose various ridge parameter estimators for this model. We derive the theoretical properties of the ridge estimator and compare its performance with the quasi-likelihood estimator in terms of matrix and scalar mean squared error. We further compared the proposed estimator numerically through a Monte Carlo simulation study and a real-life application. We found that both the simulation and application results show the superiority of the ridge estimator, particularly with the best ridge parameter estimator, over the quasi-likelihood estimator in the presence of multicollinearity issue.

A NONLINEAR REGRESSION MODEL FOR EARLY LOC ESTIMATION OF OPEN-SOURCE KOTLIN-BASED APPLICATIONS

Context. The early lines of code (LOC) estimation in software projects holds significant importance, as it directly influences the prediction of development effort, covering a spectrum of different programming languages, and open-source Kotlin-based applications in particular. The object of the study is the process of early LOC estimation of open-source Kotlin-based apps. The subject of the study is the nonlinear regression models for early LOC estimation of open-source Kotlin-based apps. Objective. The goal of the work is to build the nonlinear regression model with three predictors for early LOC estimation of open-source Kotlin-based apps based on the Box-Cox four-variate normalizing transformation to increase the confidence in early LOC estimation of these apps. Method. For early LOC estimation in open-source Kotlin-based apps, the model, confidence, and prediction intervals of nonlinear regression were constructed using the Box-Cox four-variate normalizing transformation and specialized techniques. These techniques, relying on multiple nonlinear regression analyses incorporating multivariate normalizing transformations, account for the dependencies between variables in non-Gaussian data scenarios. As a result, this method tends to reduce the mean magnitude of relative error (MMRE) and narrow confidence and prediction intervals compared to models utilizing univariate normalizing transformations. Results. An analysis has been carried out to compare the constructed model with nonlinear regression models employing decimal logarithm and Box-Cox univariate transformation. Conclusions. The nonlinear regression model with three predictors for early LOC estimation of open-source Kotlin-based apps is constructed using the Box-Cox four-variate transformation. Compared to the other nonlinear regression models, this model demonstrates a larger multiple coefficient of determination, a smaller value of the MMRE, and narrower confidence and prediction intervals. The prospects for further research may include the application of other data sets to construct the nonlinear regression model for early LOC estimation of open-source Kotlin-based apps for other restrictions on predictors.

Sequential estimation for the multiple linear regression models with balanced loss functions

Sequential analysis (SA) as a sampling technique has notable advantages like smaller average sample size and reduced value of risk compared to similarly comparable fixed-sample techniques. In this study, we first propose a few models for the estimation of the regression parameters or functions of parameters under the multiple linear regression (MLR) setup using a balanced loss function (BLF). Thereafter, we obtain the expressions of risk functions and optimal fixed sample sizes for the proposed models based on the bounded risk criteria. We establish that no fixed-sample procedures can tackle these estimation problems. Hence, we propose different multistage sampling methodologies, viz. (i) two-stage sampling, (ii) three-stage sampling, (iii) purely sequential sampling, and (iv) batch sequential sampling, and corroborate the same with the detailed simulation and real data analyses.

Software Estimation in the Design Stage with Statistical Models and Machine Learning: An Empirical Study

Accurate estimation of software effort and time in the software development process is a key activity to achieve the necessary product quality. However, underestimation or overestimation of effort has become a key challenge for software development. One of the main problems is the estimation with metrics from late stages, because the product must already be finished to make estimates. In this paper, the use of statistical models and machine learning approaches for software estimation are used in early stages such as software design, and a data set is presented with metric values of design artifacts with 37 software projects. As results, models for the estimation of development time and effort are proposed and validated through leave-one-out cross-validation. Further, machine learning techniques were employed in order to compare software projects estimations. Through the statistical tests, it was proven that the errors were not statistically different with the regression models for effort estimation. However, with Random Forest the best statistical results were obtained for estimating development time.

Disability, childhood experiences of violence and associated health outcomes in refugee settlements in Uganda

BackgroundThere is limited evidence regarding the associations between disability, childhood experiences of violence, and associated health outcomes in humanitarian settings. ObjectiveWe examined the prevalence of childhood sexual, physical, and emotional violence by disability status, explored associations between childhood violence and type of disability (limitation), perpetrator types, and the negative health outcomes associated with experiencing childhood violence by disability status. Participant and SettingParticipants included 1338 females and 927 males aged 13–24 years living in refugee settings in Uganda. MethodsData were from a cross-sectional Humanitarian Violence against Children and Youth Survey (HVACS) conducted between March and April 2022 in Uganda. Analysis entailed cross-tabulation with a chi-square test and estimation of bivariate and multivariate logistic regression models. ResultsFor both females and males, the prevalence of sexual violence in childhood was higher among those with disabilities compared to those without disabilities (23.2% vs. 11.5% for females; and 15.7% vs. 7.6% for males). The odds of experiencing sexual violence were higher among females with physical limitations (self-care [AOR:2.1; 95%CI-1.0-4.3] and task performance [AOR:2.5; 95%CI = 1.3–5.2]) and males with both physical [AOR:4.4; 95%CI = 1.4–13.7] and communication [AOR:4.1; 95%CI = 1.3–12.9] limitations compared to those without such limitations. Experiencing violence and having disabilities increased the odds of reporting negative health outcomes including severe mental distress and symptoms or being diagnosed with STI among females by three times. ConclusionIn Uganda's refugee settings, the prevalence of childhood violence is higher among children and youth with disabilities compared to those without disabilities. Females with disabilities and who had experienced childhood violence were considerably more susceptible to negative health outcomes. These findings underscore the need for targeted child protection and response interventions to address the vulnerabilities of children and youth, and particularly for those with disabilities and female children.