Backward Elimination Technique Research Articles

Resolving the Paradox: How mobile money drives economic growth through financial inclusion

The primary aim of this study is to identify the most relevant financial inclusion variables influencing gross domestic product (GDP) and to choose the optimal multiple linear regression model to quantify their contribution to economic growth. Finance is essential for every economic activity and financial inclusion is closely related to societal empowerment. It is observed in different nations that financial inclusion variables have a valuable impact on economic growth. For this, secondary data were gathered for this study from the IMF (International Monetary Fund) from 2011 to 2020. Graphical presentations were used to visualize the relationship between financial inclusion variables and economic growth. The most eligible variables and the best model among a collection of models based on empirical data were chosen using a stepwise backward elimination technique using multiple linear regression and model selection criteria. From the graph, it is seen that financial inclusion variables are proportionately related to GDP and the results demonstrate that the most eligible variables of financial inclusion that affect GDP are (1) the number of registered mobile money agent outlets, (2) the number of active mobile money accounts, (3) the number of mobile money transactions, and (4) outstanding balances on active mobile money accounts. These variables have a positive impact on economic growth, and the fitted model measures the more significant contribution to economic growth. As a result, it may be summarized that financial enclosure has to be enhanced for the development of rural and border areas of Bangladesh. This study will assist policymakers, development organizations, and scholars and make knowledgeable findings that will guide long-term improvements to minimize the gap between rich and poor, leading to the improvement of the community's wealth and welfare.

Abstract 1031: Prevalence Of Erectile Dysfunction And Its Associated Factors Among Filipino Adult Patients In A Tertiary Hospital: A Cross-sectional Study

Background: Erectile Dysfunction (ED) is an important worldwide health issue that affects nearly half of men over the age of 40 years. It has a significant negative impact on the quality of life and life satisfaction of the affected individual as well as his partner. The primary objective of this study was to determine the prevalence of erectile dysfunction and its associated factors. Methods: This was a prospective and cross-sectional survey of Filipino adult patients seen at a Tertiary Hospital from April 2023 to September 2023. The instrument used was the International Index of Erectile Function (IIEF) Questionnaire. Comparison of characteristics by ED severity were performed using One Way ANOVA or Kruskal Wallis test for continuous variables, and Chi square or Fisher’s exact test for categorical variables. Multiple logistic regression analysis with backward elimination technique was utilized to identify the factors associated with ED. P values ≤0.05 were considered statistically significant. Results: A total of 106 men aged within 18-80 years, with mean age of 43.9 (±14.4) years, were included in the study. The prevalence of ED on this study was 10.4% among Filipino adult men with equal prevalence among <40 years and ≥40 years (p=0.075). Increasing waist circumference was associated with a 1.34 (OR = 1.13, p=0.004; 95% CI: 1.10-1.63) times increase in the odds of developing mild ED while hypertension was associated with an 82.677 (OR = 27.52, p=0.008; 95% CI: 3.17-2155.26) times increase in the odds of developing mild ED. Conclusion: One out of ten Filipinos have mild ED and the associated risk factors were hypertension and waist circumference. This study recommended to increase awareness among men, patients, and doctors around the country about the existence of ED problem and its consequence so that prevention can be advised and care can be adequately provided. Keywords: Erectile dysfunction, Hypertension, Waist Circumference, Prevalence, Factors Page 1 of 2

Cognitive Inflexibility Predicts Negative Symptoms Severity in Patients with First-Episode Psychosis: A 1-Year Follow-Up Study.

Negative symptoms and cognitive deficits play a major role in psychosis and significantly influence the functional outcomes of patients, particularly those with a first episode of psychosis (FEP). However, limited research has explored the predictive capacity of cognitive deficits during FEP for subsequent negative symptomatology. Drawing from the Athens FEP research study, we conducted a retrospective longitudinal study in 80 individuals with FEP. All patients were drug naive at admission. Cognitive tests were administered at 1-month and 1-year post-admission, while negative symptomatology was assessed at the same time points using PANSS by trained raters. We considered confounding factors such as age, gender, duration of untreated psychosis (DUP), treatment received, premorbid social adjustment, and premorbid IQ. Univariate regression analysis identified cognitive domains that correlated with negative symptomatology. These, along with the confounders, were incorporated into a multiple regression, with the 1-year PANSS negative scale serving as the dependent variable. Employing the backward elimination technique, we found a statistically significant inverse relationship between the categories completed in the Wisconsin card sorting test (WCST) and the 1-year PANNS negative scale (p = 0.01), beyond the associations with DUP and the 1-month PANSS negative scale. Our results suggest that cognitive flexibility, a key component of executive functions, predicts negative symptom severity one year after FEP.

Factors Associated with Antenatal Care Visit Dropout among Mothers attending Immunization Clinic

Introduction: Developing nations continue to have high maternal mortality rates. Prenatal care helps to improve the mother's and the unborn child's health. The purpose of this study was to identify the variables linked to mothers' antenatal care visit dropout. Methods: A cross-sectional study was carried out using the multi-stage sampling technique in the four urban and four rural randomly chosen municipalities of Morang district. In total, 134 postnatal mothers who skipped four focused ANC visits were chosen as study group and 266 postnatal mothers who completed four focused ANC visit from the same vaccination facilities were chosen as control group. Using a semi-structured questionnaire and the consecutive sampling approach, data were gathered through interviews. The final model was chosen by employing the backward elimination technique after doing bivariate and multivariate analysis. Results: This study found that the majority (56.0%) of cases was from the age group of 21-30 years and 35.1% of cases were janajatis. The number of factors were linked to ANC dropout, including age (OR 2.49, 95% CI 1.07-5.82), ethnicity (OR 2.29, 95% CI 1.05-4.99), education (OR 4.64, 95% CI 1.99- 10.81), income (OR 3.49, 95% CI 1.75- 6.95), and the existence of complications (OR 1.15, 95% CI 0.54-2.44). Conclusion: The risk factors for ANC dropout were age, ethnicity, education, income, and the presence of complications during previous pregnancy. It suggests that efforts should be made to spread the information about the value of four focused ANC visits.

Shifting the gaze on implementation: examining the association between the implementation of tobacco control laws and prevalence of tobacco using data from a nationally representative survey

BackgroundTobacco use and the associated health burden is a cause of concern in India and globally. Despite several tobacco control policies in place, their sub-optimal and variable implementation across Indian states has remained a concern. Studies evaluating the real-world implementation of policies such as Cigarettes and Other Tobacco Products (COTPA) or National Tobacco Control Program (NTCP) in India and its association with reductions in tobacco use are limited. In this paper, we analyse data from a nationally representative survey to examine how policy implementation is associated with the tobacco use prevalence in India.MethodsWe analysed data from the Global Adult Tobacco Survey (GATS 2016–17) India using multivariable logistic regression. The dependent variables were the use of smoked tobacco, smokeless tobacco, and tobacco in any form. The independent variables were proxies of implementation of the COTPA and the NTCP. We followed a step-wise backward elimination technique to reach the best fit models.ResultsPeople exposed to no-smoking signages had lower odds of using tobacco (OR = 0.70, p < 0.001). People exposed to second-hand smoke (OR = 1.51, p < 0.001) and tobacco product advertisements (OR = 1.23, p < 0.001) had greater odds of using tobacco. Exposure to tobacco advertisements was associated with higher odds of using smokeless tobacco (OR = 1.23, p < 0.001), and smoked (OR = 1.33, p < 0.001) forms of tobacco.ConclusionWe find significant association between the implementation of tobacco control laws/programs and tobacco use in India. Our findings highlight the potential that policy implementation holds in reducing population-level tobacco use thus drawing attention towards the implementation phase of policies. The findings have implications on prioritising enforcement of specific tobacco control measures such as smokefree laws, modifying COTPA signages to encompass all tobacco products including against smokeless tobacco use and strengthening indirect advertising restrictions. Future research could focus on developing and validating predictors specific to policy implementation to support policy evaluation efforts.

Additional predictors of the lower limit of cerebral autoregulation during cardiac surgery.

The lower limit of autoregulation (LLA) of cerebral blood flow was previously shown to vary directly with the Ambulatory Arterial Stiffness Index (AASI) redefined as 1-regression slope of DBP-versus-SBP readings invasively measured from the radial artery before the bypass. We aimed expanding the predictive capacity of the LLA with AASI by combining it with additional predictors and provide new indications whether mean arterial pressure (MAP) is above/below the LLA. In 181 patients undergoing cardiac surgery, mean (SD) age 71 (8) years), we identified from the demographic, preoperative and intraoperative characteristics independent and statistically significant 'single predictors' of the LLA (including AASI). This was achieved using multivariate linear regression with a backward-elimination technique. The single predictors combined with 1-AASI generated new multiplicative and additive composite predictors of the LLA. Indicators for the MAP-to-LLA difference (DIF) were determined using DIF-versus-predictor plots. The odds ratio (OR) for the DIF sign (Outcome = 1 for DIF≤0) and predictor-minus-median sign (Exposure = 1 for Predictor ≤ Median) were calculated using logistic regression. BMI, 1-AASI and systolic coefficient of variation were identified single predictors that correlated similarly with the LLA ( r = -0.26 to -0.27, P < 0.001). The multiplicative and additive composite predictors displayed higher correlation with LLA ( r = -0.41 and r = -0.43, respectively, P < 0.001) and improved LLA estimation. The adjusted OR for the composite predictors was nearly twice that of the single predictors. The novel composite predictors may enhance the LLA estimation and the ability to maintain MAP in the cerebral autoregulatory range during cardiac surgery.

Effects of Leaf Wetness Duration, Temperature, and Host Phenological Stage on Infection of Walnut by Xanthomonas arboricola pv. juglandis.

Bacterial blight, caused by Xanthomonas arboricola pv. juglandis, is a significant disease affecting walnut production worldwide. Outbreaks are most severe in spring, and closely tied to host phenology and weather conditions. Pathogen infections are mainly observed in catkins, female flowers, leaves, and fruits. In this study, the effect of wetness duration and temperature on walnut infections by X. arboricola pv. juglandis was determined through two independent experiments conducted under controlled environmental conditions. The combined effect of both climatic parameters on disease severity was quantified using a third-order polynomial equation. The model obtained by linear regression and backward elimination technique fitted well to the data (R2 = 0.94 and R2adj = 0.93). The predictive capacity of the forecasting model was evaluated on pathogen-inoculated walnut plants exposed to different wetness duration-temperature combinations under Mediterranean field conditions. Observed disease severity in all events aligned with predicted infection risk. Additionally, the relationship between leaf and fruit age and the disease severity was quantified and modelled. A prediction model, which has been referred to as the WalBlight-risk model, is proposed for evaluation as an advisory system for timing bactericide sprays to manage bacterial blight in Mediterranean walnut orchards.

Iron intake during pregnancy reduces low birth weight: A propensity score analysis

Background: Low birth weight (LBW) remains a significant public health problem. As nutritional problems, including anaemia, can cause LBW, iron supplementation is needed during pregnancy. Objectives: To assess the impact of iron supplementation during pregnancy on birth weight in Indonesia using data from the 2017 Indonesia Demographic Health Survey (IDHS). Method: Birth weight was categorized into two groups, LBW (&lt;2500g) and normal birth weight (2500g or more). Iron supplementation was constructed into two groups, mothers who did not take or whose intake was less than 90 tablets during pregnancy and mothers who took ≥90 tablets during pregnancy. A sample of 10,518 children born within five years was matched using propensity scores. We used binary logistic regression and adjusted for 12 possible confounders after matching the data. Then, we performed logistic regression to assess the relationship between iron intake during pregnancy and birth weight. We built multistage multivariate logistic regression models using the backward elimination technique. We adjusted the model with some confounding factors, including socioeconomic and community-level factors, maternal and child characteristics, and maternal and child health services. Results: After adjusting for potential confounders, this study found that iron supplementation during pregnancy significantly reduced LBW (AOR: 0.75; 95% CI: 0.64 to 0.88). Conclusions: This study found that iron supplementation during pregnancy significantly reduced LBW Sri Lanka Journal of Child Health, 2023; 52(2): 136-141

188 Comparison of Nutritive Value Parameters to Normalized Difference Vegetation Index by Means of Unmanned Aerial Vehicle

Abstract Remote sensing has opened doors for grazing management by allowing producers to monitor forage in a more efficient way. The objective of this study was to observe the relationship between the normalized vegetation index (NDVI) from images taken by an unmanned aerial vehicle (UAV) to nutritive value parameters. This project took place at the H. H. Leveck Animal Research Center located in Starkville, MS where 9, 2-hectare pastures planted in Marshall annual ryegrass (Lolium multiflorum) were utilized. Each 2-hectare pasture was subdivided into 0.20-hectare sections where forage samples were collected from 3, 1/4 of a meter quadrat every 14-days. Images were taken with UAV over a 76-day grazing period every 14-days. Images from the UAV were captured at 121.9 m above ground level using a DJI inspire 2 rotor wing fitted with a MicaSense RedEdge camera (MicaSense; Seattle, WA). Images were processed using Pix4D (Pix4D SA; Prilly, Switzerland) and analyzed in QGIS (QGIS; OSGeo; Finland). Raster calculator in QGIS was utilized to calculate NDVI values. Forage samples were analyzed using NIR for crude protein (CP), acid detergent fiber (ADF), neutral detergent fiber (NDF), relative forage quality (RFQ), and invitro dry matter digestibility (IVDMD) and correlated to NDVI values. Regression analysis was conducted in R (RStudio, PBC) using the package easynls (easynls-package) procedure within package ggplot2 (ggplot2-package). Regression models were obtained through a backward-elimination technique. Neutral detergent fiber (r2 = 0.27) had the best correlated relationship to NDVI compared with CP (r2 = 0.21), ADF (r2 = 0.25), RFQ (r2 = 0.18), and IVDMD (r2 = 0.22). As NDVI values increased, CP, RFQ, and IVDMD increased while ADF and NDF decreased. There was a weak correlation between all nutritive value parameters and NDVI values. In conclusion, weak relationships indicate that NDVI values may not be the best predictor of nutritive value.

The Efficiency of Seven-Variable Box-Behnken Experimental Design with Varying Center Runs on Full and Reduced Model Types

Response Surface Methodology is widely used in the optimization of industrial processes and products that depend on several experimental variables. One of the most tested and efficient second-order Response Surface Methodology designs is the Box-Behnken Design. This research explores the efficiency of a seven-variable Box-Behnken design with varying center runs, on full and reduced quadratic models. Backward Elimination and Forward Selection techniques are employed as the variable selection techniques for obtaining the reduced models, based on the Akaike Information Criterion. Fit Statistics and Design Efficiency values are obtained for the reduced models and are compared with those of the full model. Generally, results show that the reduced quadratic models perform best under one center-point run, thereby making the reduced models the most preferred in terms of the model fit and D-efficiency. Comparative analysis, based on G-efficiency, reveals that the full quadratic model performs better than the reduced models under one center-point Box-Behnken design.

Evaluation and Multi-Objective Optimization of Lightweight Mortars Parameters at Elevated Temperature via Box-Behnken Optimization Approach.

In this research, the mechanical properties of lightweight mortars containing different percentages of additional powder materials has been investigated using response surface methodology (RSM). Box–Behnken design, one of the RSM techniques, was used to study the effects of silica fume content (5, 10, and 15%), vermiculite/cement (V/C) ratio (4, 6, and 8), and temperature (300, 600, and 900 °C) on the ultrasonic pulse velocity (UPV), bending strength, and compressive strength of lightweight mortars. Design expert statistical software was accustomed to determining and evaluating the mix-design of materials in mortar mixtures and temperature effect on mortars. After preliminary experimental research of the relationships between independent and response variables, regression models were built. During the selection of the model parameters, F value, p-value, and R2 values of the statistical models were taken into account by using the backward elimination technique. The results showed a high correlation between the variables and responses. Multi-objective optimization results showed that the critical temperatures for different levels of silica fume (5–10–15%) were obtained as 371.6 °C, 306.3 °C, and 436 °C, respectively, when the V/C ratio kept constant as 4. According to the results obtained at high desirability levels, it is found that the UPS values varied in the range of 2480–2737 m/s, flexural strength of 3.13–3.81 MPa, and compressive strength of 9.9–11.5 MPa at these critical temperatures. As a result of this research, RSM is highly recommended to evaluate mechanical properties where concrete includes some additional powder materials and was exposed to high temperature.

Bayesian modeling and forecasting of vector autoregressive moving average processes

The article proposes a Bayesian methodology to implement complete and cohesive analysis of vector time series. Assuming the series are generated by vector autoregressive moving average process, the identification, diagnostic checking, estimation and forecasting phases of time series analysis are done by referring to the appropriate posterior or predictive distributions. The identification phase is based on approximating the posterior distribution of the coefficients of the largest possible model by a matric-variate generalization of the t-distribution (matrix t-distribution). Then the insignificant coefficients are eliminated by a sequence of F or Chi-square tests using a procedure similar to the backward elimination technique used in regression analysis. The diagnostic checking phase is done using overfitting tests, the estimation phase is done using the matrix t and Wishart distributions, and the forecasting phase is performed using multivariate t-distribution. Using the proposed posterior distributions, the article proposes the machinery necessary to implement the four phases of multivariate time series analysis. The proposed Bayesian methodology has been illustrated and checked by a simulated data used by distinguished researchers. The Initial examination of the numerical results supports the adequacy of using the proposed methodology.

Influence of the characteristics of home office work on self-perceived vocal fatigue during the COVID-19 pandemic

Purpose To verify the influence of work characteristics and the occupational voice use on the self-perception of vocal fatigue symptoms in individuals working in the home office during the COVID-19 pandemic. Methods This was a cross-sectional, observational, and descriptive study. The sample consisted of 206 individuals (123 women and 83 men), with an average age of 34 years, working exclusively in a home office mode due to the pandemic. Through an online form, all participants responded to the Vocal Fatigue Index (VFI) protocol and the questionnaire on the characterization of work and the occupational voice use during the pandemic, developed by the authors of this study. Multiple linear regression using the backward elimination technique was performed. Results The variables the interlocutor does “not listen to me in home office work,” “noise in the home office work environment,” “the daily workload in home office,” “vocal quality worsened in home office work,” “increased vocal loudness in home office work” and “lack of training about voice use in home office work” are predictors of the dependent variable vocal fatigue symptoms. Conclusion The characteristics of work and occupational voice use influence the self-perception of vocal fatigue symptoms in individuals working in the home office mode during the COVID-19 pandemic.

A Cross-sectional Study of Perceptual and Acoustic Voice Characteristics in Healthy Aging

The human voice qualitatively changes across the lifespan. Although some of these vocal changes may be pathologic, other changes likely reflect natural physiological aging. Normative data for voice characteristics in healthy aging is limited and disparate studies have used a range of different acoustic features, some of which are implicated in pathologic voice changes. We examined the perceptual and acoustic features that predict healthy aging. Participants (N = 150) aged between 50 and 92 years performed a sustained vowel task. Acoustic features were measured using the Multi-Dimensional Voice Program and the Analysis of Dysphonia in Speech and Voice. We used forward and backward variable elimination techniques based on the Bayesian information criterion and linear regression to assess which of these acoustic features predict age and perceptual features. Hearing thresholds were determined using pure-tone audiometry tests at frequencies 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, and 4000 Hz. We further explored potential relationships between these acoustic features and clinical assessments of voice quality using the Consensus Auditory-Perceptual Evaluation of Voice. Chronological age was significantly predicted by greater voice turbulence, variability of cepstral fundamental frequency, low relative to high spectral energy, and cepstral intensity. When controlling for hearing loss, age was significantly predicted by amplitude perturbations and cepstral intensity. Clinical assessments of voice indicated perceptual characteristics of speech were predicted by different acoustic features. For example, breathiness was predicted by the soft phonation index, mean cepstral peak prominence, mean low-high spectral ratio, and mean cepstral intensity. Findings suggest that acoustic features that predict healthy aging are different than those previously reported for the pathologic voice. We propose a model of healthy and pathologic voice development in which voice characteristics are mediated by the inability to monitor vocal productions associated with age-related hearing loss. This normative data of healthy vocal aging may assist in separating voice pathologies from healthy aging.

Predicting Heart Rate Variability Parameters in Healthy Korean Adults: A Preliminary Study.

The purpose of the study was to examine the development of a multiple linear regression model to estimate heart rate variability (HRV) parameters using easy-to-measure independent variables in preliminary experiments. HRV parameters (time domain: SDNN, RMSSD, NN50, pNN50; frequency domain: TP, VLF, LF, HF) and the independent variables (e.g., sex, age, body height, body weight, BMI, HR, HRmax, HRR) were measured in 75 healthy adults (male n = 27, female n = 48) for estimating HRV. The HRV estimation multiple linear regression model was developed using the backward elimination technique. The regression model’s coefficient of determination for the time domain variables was significantly high (SDNN = R2: 72.2%, adjusted R2: 69.8%, P < .001; RMSSD = R2: 93.1%, adjusted R2: 92.1%, P < .001; NN50 = R2: 78.0%, adjusted R2: 74.9%, P < .001; pNN50 = R2: 89.1%, adjusted R2: 87.4%, P < .001). The coefficient of determination of the regression model for the frequency domain variable was moderate (TP = R2: 75.6%, adjusted R2: 72.6%, P < .001; VLF = R2: 41.6%, adjusted R2: 40.3%, P < .001; LF = R2: 54.6%, adjusted R2: 49.2%, P < .001; HF = R2: 67.5%, adjusted R2: 63.4%, P < .001). The coefficient of determination of time domain variables in the developed multiple regression models was shown to be very high (adjusted R2: 69.8%–92.1%, P < .001), but the coefficient of determination of frequency domain variables was moderate (adjusted R2: 40.3%–72.6%, P < .001). In addition to the equipment used for measuring HRV in clinical trials, this study confirmed that simple physiological variables could predict HRV.

Machine Learning Modelling and Feature Engineering in Seismology Experiment.

This article aims to discusses machine learning modelling using a dataset provided by the LANL (Los Alamos National Laboratory) earthquake prediction competition hosted by Kaggle. The data were obtained from a laboratory stick-slip friction experiment that mimics real earthquakes. Digitized acoustic signals were recorded against time to failure of a granular layer compressed between steel plates. In this work, machine learning was employed to develop models that could predict earthquakes. The aim is to highlight the importance and potential applicability of machine learning in seismology The XGBoost algorithm was used for modelling combined with 6-fold cross-validation and the mean absolute error (MAE) metric for model quality estimation. The backward feature elimination technique was used followed by the forward feature construction approach to find the best combination of features. The advantage of this feature engineering method is that it enables the best subset to be found from a relatively large set of features in a relatively short time. It was confirmed that the proper combination of statistical characteristics describing acoustic data can be used for effective prediction of time to failure. Additionally, statistical features based on the autocorrelation of acoustic data can also be used for further improvement of model quality. A total of 48 statistical features were considered. The best subset was determined as having 10 features. Its corresponding MAE was 1.913 s, which was stable to the third decimal point. The presented results can be used to develop artificial intelligence algorithms devoted to earthquake prediction.

Ipilimumab (IPI) alone or in combination with anti-PD-1 (IPI+PD1) in patients (pts) with metastatic melanoma (MM) resistant to PD1 monotherapy.

10005 Background: PD1 induces long-term responses in approximately 30% of MM pts, however 2/3 are resistant (innate or acquired) and will require further treatment. A subset of these pts will benefit from IPI or IPI+PD1, but these pts are yet to be identified. We sought to determine; i) response rate (RR) and survival to IPI+/-PD1 after PD1 progression, and ii) clinical predictors of response and survival to IPI+/-PD1. Methods: MM pts resistant to PD1 and then treated with IPI+/-PD1 were studied. Demographics, disease characteristics and baseline blood parameters were examined. Univariate, multivariate and backward elimination technique analyses were performed to create predictive models of response and overall survival (OS). Results: Of 330 MM pts resistant to PD1 (median time to prog 2.9 months [0.5 – 42.3], 12% adjuvant, 88% metastatic; 70% innate, 30% acquired), 161 (49%) had subsequent IPI and 169 (51%) had IPI+PD1. Characteristics at start of IPI+/-PD1 were similar in IPI vs IPI+PD1 groups (stage M1D 27% vs 34%; elevated LDH 38% vs 40%), except IPI group had more ECOG ≥1 (60% vs 34%) and less BRAF mutation (mut) (21% vs 37%). Median follow-up from start of IPI+/-PD1 was 22.3 months (19.8 - 25.8); RR was 22%, higher in IPI+PD1 (31%) vs IPI (12%) (p &lt; 0.01). PFS and OS at 1 year were 20% and 48%, respectively; better with IPI+PD1 (27%/57%) vs IPI (13%/38%) (p &lt; 0.01). PD1 setting (adjuvant/metastatic) and response did not impact response to IPI+/-PD1. Most pts progressing on adjuvant PD1 had IPI+PD1 (88%) and RR was 33%. Neither the interval between PD1 and IPI+/-PD1 nor use of other drugs affected response to IPI+/-PD1. RR was similar in BRAF WT (23%) vs BRAF mut (RR 21%) pts. In BRAF WT pts, RR was higher with IPI+PD1 vs IPI (38% vs 9%, p &lt; 0.01), while RR was similar with IPI (24%) or IPI+PD1 (19%) in BRAF mut pts. One third of BRAF mut pts had BRAF inhibitors (BRAFi) prior to IPI+/-PD1 and lower RR (13%) vs those without BRAFi (RR = 25%, p &gt; 0.05). High grade (≥G3) toxicity (tox) was similar with IPI+PD1 (30%) or IPI (34%, p = 0.48), and was not associated with response. Stage III/M1A/M1B, normal LDH and treatment with IPI+PD1 were the best predictors of response (AUC = 0.69). These factors, in addition to sex (male), ECOG PS = 0, BRAF mut, progressed/recurred &gt; 3 months on PD1, and absence of bone mets were the best predictors of longer OS (AUC = 0.74). Conclusions: In pts resistant to PD1, IPI+PD1 has higher RR, longer survival, yet similar high grade tox than IPI alone. Predictive models of response &amp; survival will help select pts for IPI+/-PD1 after progressing on PD1.

Intimate Partner Violence Among Women Accessing HIV-Care Services at Tertiary Hospitals in Ogun State, Nigeria: Implications for Policy and Practice.

Intimate partner violence is both a predisposing factor for and sequelae of HIV infection. It is more likely for HIV-positive women to experience intimate partner violence compared with HIV-negative women. Previous researches have described the other risk factors for intimate partner violence. This cross-sectional study identified the correlates of intimate partner violence among HIV-positive women accessing HIV-care. We interviewed 458 women accessing HIV-care at the three tertiary hospitals in Ogun State, Nigeria. Trained research assistants conducted 10-min structured interviews with validated questionnaires. We carried out descriptive, bivariate, and regression analyses. We used the backward elimination technique to build a model to predict the experience of intimate partner violence (IPV) within the preceding 12 months. We set the level of significance at .05. A total of 23 (5.02%) participants reported IPV within the preceding 12 months, while the prevalence of lifetime intimate partner violence was 24.02%. Younger age, the experience of IPV in previous relationships, and having multiple sex partners were related to the experience of IPV within the preceding 12 months (p < .05). Also, the partners' age, alcohol intake, and current smoking status were associated with the experience of IPV within the preceding 12 months (p < .05). After regression analysis, participants' age (adjusted odds ratio [AOR] = 0.892, 95% confidence interval [CI] = [0.831, 0.957]), experience of IPV in previous relationship (AOR = 12.841, 95% CI = [4.303, 38.318]), and partners' current smoking status (AOR = 4.874, 95% CI = [1.252, 18.969]) retained association with the experience of IPV within the preceding 12 months. IPV among HIV-positive women accessing HIV-care occurs in a complicated, context-specific way. Routine HIV strategies and services should include interventions that involve men and address gender power imbalances. For effectiveness, specific interventions must consider women's' specific context and characteristics.

Backward Eliminated Formulation of Fire Area Coverage using Machine Learning Regression

In today’s modern world, the environmental wealth is degraded due to the advancement in the technology. The software development leads to the emission of electronic waste that affects the whole part of the country. The forest area and the agricultural land are converted into living places, companies, industries, server warehouse and heavy workstations. Due to this, heavy damages occur in the environmental resources. The basic characteristics of the nature quality are becoming poor due to the technological advancement. Due to the heavy emission of rays in the environment, there is a chance for the occurrence of fire in the forest. This leads to the challenging issue of predicting the area coverage of the fire in the forest. After the event of fire damage, it is a difficult task to analyze the area that suffered from the fire. With this analytical view, this paper focuses on finding the area coverage of fire using various regression algorithms. The forest area coverage dataset from the UCI machine learning repository is used for analyzing the area coverage of the fire. The prediction of area coverage of fire is accomplished in four ways. Firstly, the raw data set is fitted with various regression algorithms to predict the fire area coverage. Secondly, the data set is tailored by the feature selection algorithm namely backward elimination technique. Thirdly, the backward eliminated reduced fire area coverage data set is fitted with various regression algorithms to predict the fire area coverage. Fourth, the performance analysis is done for the raw data set and backward eliminated reduced fire area coverage data set by reviewing the performance metrics mean squared error (MSE), Mean Absolute Error (MAE) and R2 Score. This paper is implemented by python scripts in Anaconda Spyder Navigator. Experimental Result shows that the Passive Aggressive regressor have the effective prediction of fire area coverage with minimum MSE of 0.07, MAE of 1.03 and equitable R2 Score of 0.93 without backward elimination. In the same way, the Passive Aggressive regressor MSE of 0.06, MAE of 1.02 and equitable R2 Score of 0.96 with backward elimination.

Comparative study on the effects of meteorological and pollutant parameters on ANN modelling for prediction of SO2

Variations in meteorological parameters, different transportation mechanisms, complex reaction mechanisms and insufficient control measures have made control of ambient air pollution, a challenge. Prediction mechanisms for pollutant concentrations in advance have become necessary to regulate parameters within the acceptable limits. Artificial neural network (ANN) modelling can be used for predicting the concentrations of pollutants by establishing functional relationships between complex and nonlinear predictor variables and outputs. The application of time-series ANN models to predict air quality parameters was investigated by using six pollutant variables, five meteorological parameters and three time parameters to predict the concentration of SO2. An industrial belt in the southern part of India was selected as the study area. Two years of input parameters and feed-forward back propagation algorithm were used to construct the ANN model. Input parameters were optimized using forward selection and backward elimination techniques. Mean squared error and coefficient of determination were used to evaluate the models. The developed model exhibited very promising performance evaluation characteristics with the best model resulting in MSE of 0.0115 and R2 value of 0.8979. The hybrid ANN models employing input parameter optimization techniques resulted in better performance characteristic values than conventional models. Any model exhibited a minimum reduction in MSE by 9% and 3% improvement in correlation. Pollutant parameters were found to affect the ANN models when compared to the meteorological parameters. Predicted values compared to the National Ambient Air Quality Standards were as low as 55% of the maximum allowable concentration in ambient air.