Search results for: penalized logistic regression

Authors: Rohini Hariharan, Yazhini R., Blessy Maria Mathew

Abstract:

The task of detecting email spam is a very important one in the era of digital technology that needs effective ways of curbing unwanted messages. This paper presents an approach aimed at making email spam categorization algorithms transparent, reliable and more trustworthy by incorporating Local Interpretable Model-agnostic Explanations (LIME). Our technique assists in providing interpretable explanations for specific classifications of emails to help users understand the decision-making process by the model. In this study, we developed a complete pipeline that incorporates LIME into the spam classification framework and allows creating simplified, interpretable models tailored to individual emails. LIME identifies influential terms, pointing out key elements that drive classification results, thus reducing opacity inherent in conventional machine learning models. Additionally, we suggest a visualization scheme for displaying keywords that will improve understanding of categorization decisions by users. We test our method on a diverse email dataset and compare its performance with various baseline models, such as Gaussian Naive Bayes, Multinomial Naive Bayes, Bernoulli Naive Bayes, Support Vector Classifier, K-Nearest Neighbors, Decision Tree, and Logistic Regression. Our testing results show that our model surpasses all other models, achieving an accuracy of 96.59% and a precision of 99.12%.

Keywords: text classification, LIME (local interpretable model-agnostic explanations), stemming, tokenization, logistic regression.

Procedia PDF Downloads 48

Authors: Suzan Arslanturk, Mohammad-Reza Siadat, Theophilus Ogunyemi, Ananias Diokno

Abstract:

Missing data is a common challenge in statistical analyses of most clinical survey datasets. A variety of methods have been developed to enable analysis of survey data to deal with missing values. Imputation is the most commonly used among the above methods. However, in order to minimize the bias introduced due to imputation, one must choose the right imputation technique and apply it to the correct type of missing data. In this paper, we have identified different types of missing values: missing data due to skip pattern (SPMD), undetermined missing data (UMD), and genuine missing data (GMD) and applied rough set imputation on only the GMD portion of the missing data. We have used rough set imputation to evaluate the effect of such imputation on prediction by generating several simulation datasets based on an existing epidemiological dataset (MESA). To measure how well each dataset lends itself to the prediction model (logistic regression), we have used p-values from the Wald test. To evaluate the accuracy of the prediction, we have considered the width of 95% confidence interval for the probability of incontinence. Both imputed and non-imputed simulation datasets were fit to the prediction model, and they both turned out to be significant (p-value < 0.05). However, the Wald score shows a better fit for the imputed compared to non-imputed datasets (28.7 vs. 23.4). The average confidence interval width was decreased by 10.4% when the imputed dataset was used, meaning higher precision. The results show that using the rough set method for missing data imputation on GMD data improve the predictive capability of the logistic regression. Further studies are required to generalize this conclusion to other clinical survey datasets.

Keywords: rough set, imputation, clinical survey data simulation, genuine missing data, predictive index

Procedia PDF Downloads 169

Authors: Yinglisong, Jingge Chen, Jingxuan Chen, Yan Wang, Hui Huang, Jing Zhnag, Qianqian Zhang, Zhenzhen Zhang, Ji Zhang

Abstract:

Purpose: The study aims to establish a nomogram prediction model for postpartum hemorrhage (PPH) in vaginal delivery. Patients and Methods: Clinical data were retrospectively collected from vaginal delivery patients admitted to a hospital in Zhengzhou, China, from June 1, 2022 - October 31, 2022. Univariate and multivariate logistic regression were used to filter out independent risk factors. A nomogram model was established for PPH in vaginal delivery based on the risk factors coefficient. Bootstrapping was used for internal validation. To assess discrimination and calibration, receiver operator characteristics (ROC) and calibration curves were generated in the derivation and validation groups. Results: A total of 1340 cases of vaginal delivery were enrolled, with 81 (6.04%) having PPH. Logistic regression indicated that history of uterine surgery, induction of labor, duration of first labor, neonatal weight, WBC value (during the first stage of labor), and cervical lacerations were all independent risk factors of hemorrhage (P <0.05). The area-under-curve (AUC) of ROC curves of the derivation group and the validation group were 0.817 and 0.821, respectively, indicating good discrimination. Two calibration curves showed that nomogram prediction and practical results were highly consistent (P = 0.105, P = 0.113). Conclusion: The developed individualized risk prediction nomogram model can assist midwives in recognizing and diagnosing high-risk groups of PPH and initiating early warning to reduce PPH incidence.

Keywords: vaginal delivery, postpartum hemorrhage, risk factor, nomogram

Procedia PDF Downloads 79

Authors: Uswah Ahmad Khan, Muhammad Moazam Fraz, Humayoon Shafique Satti, Qasim Aziz

Abstract:

Approximately 10-14% of the global population experiences a functional disorder known as irritable bowel syndrome (IBS). The disorder is defined by persistent abdominal pain and an irregular bowel pattern. IBS significantly impairs work productivity and disrupts patients' daily lives and activities. Although IBS is widespread, there is still an incomplete understanding of its underlying pathophysiology. This study aims to help characterize the phenotype of IBS patients by differentiating the comorbidities found in IBS patients from those in non-IBS patients using machine learning algorithms. In this study, we extracted samples coding for IBS from the UK BioBank cohort and randomly selected patients without a code for IBS to create a total sample size of 18,000. We selected the codes for comorbidities of these cases from 2 years before and after their IBS diagnosis and compared them to the comorbidities in the non-IBS cohort. Machine learning models, including Decision Trees, Gradient Boosting, Support Vector Machine (SVM), AdaBoost, Logistic Regression, and XGBoost, were employed to assess their accuracy in predicting IBS. The most accurate model was then chosen to identify the features associated with IBS. In our case, we used XGBoost feature importance as a feature selection method. We applied different models to the top 10% of features, which numbered 50. Gradient Boosting, Logistic Regression and XGBoost algorithms yielded a diagnosis of IBS with an optimal accuracy of 71.08%, 71.427%, and 71.53%, respectively. Among the comorbidities most closely associated with IBS included gut diseases (Haemorrhoids, diverticular diseases), atopic conditions(asthma), and psychiatric comorbidities (depressive episodes or disorder, anxiety). This finding emphasizes the need for a comprehensive approach when evaluating the phenotype of IBS, suggesting the possibility of identifying new subsets of IBS rather than relying solely on the conventional classification based on stool type. Additionally, our study demonstrates the potential of machine learning algorithms in predicting the development of IBS based on comorbidities, which may enhance diagnosis and facilitate better management of modifiable risk factors for IBS. Further research is necessary to confirm our findings and establish cause and effect. Alternative feature selection methods and even larger and more diverse datasets may lead to more accurate classification models. Despite these limitations, our findings highlight the effectiveness of Logistic Regression and XGBoost in predicting IBS diagnosis.

Keywords: comorbidities, disease association, irritable bowel syndrome (IBS), predictive analytics

Procedia PDF Downloads 119

Authors: Amirmohsen Behjat, Aleck Ostry, Christina Miewald, Bernie Pauly

Abstract:

Farmers’ Markets have become one of the important healthy food suppliers in both rural communities and urban settings. Farmers’ markets are evolving and their number has rapidly increased in the past decade. Despite this drastic increase, the distribution of the farmers’ markets is not even across different areas. The main goal of this study is to explore the socioeconomic, geographic, and demographic variables which affect the establishment of farmers’ market in rural communities in British Columbia (BC). Thus, the data on available farmers’ markets in rural areas were collected from BC Association of Farmers’ Markets and spatially joined to BC map at Dissemination Area (DA) level using ArcGIS software to link the farmers’ market to the respective communities that they serve. Then, in order to investigate this issue and understand which rural communities farmer’ markets tend to operate, a binary logistic regression analysis was performed with the availability of farmer’ markets at DA-level as dependent variable and Deprivation Index (DI), Metro Influence Zone (MIZ) and population as independent variables. The results indicated that DI and MIZ variables are not statistically significant whereas the population is the only which had a significant contribution in predicting the availability of farmers’ markets in rural BC. Moreover, this study found that farmers’ markets usually do not operate in rural food deserts where other healthy food providers such as supermarkets and grocery stores are non-existent. In conclusion, the presence of farmers markets is not associated with socioeconomic and geographic characteristics of rural communities in BC, but farmers’ markets tend to operate in more populated rural communities in BC.

Keywords: farmers’ markets, socioeconomic and demographic variables, metro influence zone, logistic regression, ArcGIS

Procedia PDF Downloads 189

Authors: Lazaro E. Nnko, Japhet J. Kashaigili, Gerald C. Monela, Pantaleo K. T. Munishi

Abstract:

Agroforestry systems and practices are perceived to improve livelihood and sustainable management of natural resources. However, their adoption in various regions differs with the biophysical conditions and societal characteristics. This study was conducted in Kilombero District to investigate the factors influencing the adoption of different agroforestry systems and practices in agro-ecosystems and farming systems. A household survey, key informant interviews, and focus group discussion was used for data collection in three villages. Descriptive statistics and multinomial logistic regression in SPSS were applied for analysis. Results show that Igima and Ngajengwa villages had home garden practices dominated, as revealed by 63.3% and 66.7%, respectively, while Mbingu village had mixed intercropping practice with 56.67%. Agrosilvopasture systems were dominant in Igima and Ngajengwa villages with 56.7% and 66.7%, respectively, while in Mbingu village, the dominant system was agrosilviculture with 66.7%. The results from multinomial logistic regression show that different explanatory variable was statistical significance as predictors of the adoption of agroforestry systems and practices. Residence type and sex were the most dominant factor influencing the adoption of agroforestry systems. Duration of stay in the village, availability of extension education, residence, and sex were the dominant factor influencing the adoption of agroforestry practices. The most important and statistically significant factors among these were residence type and sex. The study concludes that agroforestry will be more successful if the local priorities, which include social-economic need characteristics of the society, will be considered in designing systems and practices. The socio-economic need of the community should be addressed in the process of expanding the adoption of agroforestry systems and practices.

Keywords: agroforestry adoption, agroforestry systems, agroforestry practices, agroforestry, Kilombero

Procedia PDF Downloads 118

Authors: Iyanujesu Adereti, Tajudeen Basiru, Ayodamola Olanipekun

Abstract:

Background: Substance abuse among adolescents is of public health concern globally. Despite being the most abused by adolescents, there are limited studies on the prevalence of alcohol use and cigarette smoking among adolescents in Mauritius. Objectives: To determine the prevalence of cigarette smoking, alcohol use and associated correlates among school-going adolescents in Mauritius. Methodology: Data obtained from 2017 WHO Global School-based Student Health Survey (GSHS) survey of 3,012 school-going adolescents in Mauritius was analyzed using STATA. Descriptive statistics were used to obtain prevalence. Bivariate and multivariate logistic regression analysis was used to evaluate predictors of cigarette smoking and alcohol use. Results: Prevalence of alcohol consumption and cigarette smoking were 26.0% and 17.1%, respectively. Smoking and alcohol use was more prevalent among males, younger adolescents, and those in higher school grades (p-value <.000). In multivariable logistic regression, male gender was associated with a higher risk of cigarette smoking (adjusted Odds Ratio (aOR) [95%Confidence Interval (CI)]= 1.51[1.06-2.14]) but lower risk of alcohol use (aOR[95%CI]= 0.69[0.53-0.90]) while older age (mid and late adolescence) and parental smoking were found to be associated with increased risk of alcohol use (aOR[95%CI]= 1.94[1.34-2.99] and 1.36[1.05-1.78] respectively). Marijuana use, truancy, being in a fight and suicide ideation were associated with increased odds of alcohol use (aOR[95%CI]= 3.82[3.39-6.09]; 2.15[1.62-2.87]; 1.83[1.34-2.49] and 1.93[1.38-2.69] respectively) and cigarette smoking (aOR[95%CI]= 17.28[10.4 - 28.51]; 1.73[1.21-2. 49]; 1.67[1.14-2.45] and 2.17[1.43-3.28] respectively) while involvement in sexual activity was associated with reduced risk of alcohol use (aOR[95%CI]= 0.50[0.37-0.68]) and cigarette smoking (aOR[95%CI]= 0.47[0.33-0.69]). Parental support and parental monitoring were uniquely associated with lower risk of cigarette smoking (aOR[95%CI]= 0.69[0.47-0.99] and 0.62[0.43-0.91] respectively). Conclusion: The high prevalence of alcohol use and cigarette smoking in this study shows the need for the government of Mauritius to enhance policies that will help address this issue putting into accounts the various risk and protective factors.

Keywords: adolescent health, alcohol use, cigarette smoking, global school-based student health survey

Procedia PDF Downloads 253

Authors: Roselyn Chipojola, Shu-yu Kuo

Abstract:

Background: Breastfeeding self-efficacy in both mothers and fathers play a crucial role in improving exclusive breastfeeding rates. However, less is known on the relationship and predictors of paternal and maternal breastfeeding self-efficacy. This study aimed to examine the relationship and associated factors of breastfeeding self-efficacy (BSE) among mothers and fathers in Malawi. Methods: A cross-sectional study was conducted on 180 pairs of postpartum mothers and fathers at a tertiary maternity facility in central Malawi. BSE was measured using the Breastfeeding Self-Efficacy Scale Short-Form. Depressive symptoms were assessed by the Edinburgh Postnatal Depression Scale. A structured questionnaire was used to collect demographic and health variables. Data were analyzed using multivariable logistic regression and multinomial logistic regression. Results: A higher score of self-efficacy was found in mothers (mean=55.7, Standard Deviation (SD) =6.5) compared to fathers (mean=50.2, SD=11.9). A significant association between paternal and maternal breastfeeding self-efficacy was found (r= 0. 32). Age, employment status, mode of birth was significantly related to maternal and paternal BSE, respectively. Older age and caesarean section delivery were significant factors of combined BSE scores in couples. A higher BSE score in either the mother or her partner predicted higher exclusive breastfeeding rates. BSE scores were lower when couples’ depressive symptoms were high. Conclusion: BSE are highly correlated between Malawian mothers and fathers, with a relatively higher score in maternal BSE. Importantly, a high BSE in couples predicted higher odds of exclusive breastfeeding, which highlights the need to include both mothers and fathers in future breastfeeding promotion strategies.

Keywords: paternal, maternal, exclusive breastfeeding, breastfeeding self‑efficacy, malawi

Procedia PDF Downloads 70

Authors: Vineet Agrawal, Deepanjali S., Medha R., Subitha L.

Abstract:

Background. Hospital malnutrition is a highly prevalent issue and is known to increase the morbidity, mortality, length of hospital stay, and cost of care. In India, studies on hospital malnutrition have been restricted to ICU, post-surgical, and cancer patients. We designed this study to assess the impact of hospital malnutrition on 30-day post-discharge and in-hospital mortality in patients admitted in the general medicine department, irrespective of diagnosis. Methodology. All patients aged above 18 years admitted in the medicine wards, excluding medico-legal cases, were enrolled in the study. Nutritional assessment was done within 72 h of admission, using Subjective Global Assessment (SGA), which classifies patients into three categories: Severely malnourished, Mildly/moderately malnourished, and Normal/well-nourished. Anthropometric measurements like Body Mass Index (BMI), Triceps skin-fold thickness (TSF), and Mid-upper arm circumference (MUAC) were also performed. Patients were followed-up during hospital stay and 30 days after discharge through telephonic interview, and their final diagnosis, comorbidities, and cause of death were noted. Multivariate logistic regression and cox regression model were used to determine if the nutritional status at admission independently impacted mortality at one month. Results. The prevalence of malnourishment by SGA in our study was 67.3% among 395 hospitalized patients, of which 155 patients (39.2%) were moderately malnourished, and 111 (28.1%) were severely malnourished. Of 395 patients, 61 patients (15.4%) expired, of which 30 died in the hospital, and 31 died within 1 month of discharge from hospital. On univariate analysis, malnourished patients had significantly higher morality (24.3% in 111 Cat C patients) than well-nourished patients (10.1% in 129 Cat A patients), with OR 9.17, p-value 0.007. On multivariate logistic regression, age and higher Charlson Comorbidity Index (CCI) were independently associated with mortality. Higher CCI indicates higher burden of comorbidities on admission, and the CCI in the expired patient group (mean=4.38) was significantly higher than that of the alive cohort (mean=2.85). Though malnutrition significantly contributed to higher mortality on univariate analysis, it was not an independent predictor of outcome on multivariate logistic regression. Length of hospitalisation was also longer in the malnourished group (mean= 9.4 d) compared to the well-nourished group (mean= 8.03 d) with a trend towards significance (p=0.061). None of the anthropometric measurements like BMI, MUAC, or TSF showed any association with mortality or length of hospitalisation. Inference. The results of our study highlight the issue of hospital malnutrition in medicine wards and reiterate that malnutrition contributes significantly to patient outcomes. We found that SGA performs better than anthropometric measurements in assessing under-nutrition. We are of the opinion that the heterogeneity of the study population by diagnosis was probably the primary reason why malnutrition by SGA was not found to be an independent risk factor for mortality. Strategies to identify high-risk patients at admission and treat malnutrition in the hospital and post-discharge are needed.

Keywords: hospitalization outcome, length of hospital stay, mortality, malnutrition, subjective global assessment (SGA)

Procedia PDF Downloads 150

Authors: Keiko Naito, Y. Nakashima, S. Yamauchi, Y. Kunitani, Y. Ishigami, K. Numata, M. Mizobe, Y. Homma, J. Takahashi, T. Inoue, T. Shiga, H. Funakoshi

Abstract:

Background: A presence of severe acidemia, defined as pH < 7.2, is common during endotracheal intubation for critically ill patients in the emergency department (ED). Severe acidemia is widely recognized as a predisposing factor for intubation failure. However, it is unclear that acidemic condition itself actually makes endotracheal intubation more difficult. We aimed to evaluate if a presence of severe acidemia before intubation is associated with the lower first attempt intubation success rate in the ED. Methods: This is a retrospective observational cohort study in the ED of an urban hospital in Japan. The collected data included patient demographics, such as age, sex, and body mass index, presence of one or more factors of modified LEMON criteria for predicting difficult intubation, reasons for intubation, blood gas levels, airway equipment, intubation by emergency physician or not, and the use of the rapid sequence intubation technique. Those with any of the following were excluded from the analysis: (1) no blood gas drawn before intubation, (2) cardiopulmonary arrest, and (3) under 18 years of age. The primary outcome was the first attempt intubation success rates between a severe acidemic patients (SA) group and a non-severe acidemic patients (NA) group. Logistic regression analysis was used to test the first attempt success rates for intubations between those two groups. Results: Over 5 years, a total of 486 intubations were performed; 105 in the SA group and 381 in the NA group. The univariate analysis showed that the first attempt intubation success rate was lower in the SA group than in the NA group (71.4% vs 83.5%, p < 0.01). The multivariate logistic regression analysis identified that severe acidemia was significantly associated with the first attempt intubation failure (OR 1.9, 95% CI 1.03-3.68, p = 0.04). Conclusions: A presence of severe acidemia before endotracheal intubation lowers the first attempt intubation success rate in the ED.

Keywords: acidemia, airway management, endotracheal intubation, first-attempt intubation success rate

Procedia PDF Downloads 248

Authors: Alanoud Moraya Aldalan, Abdulaziz Almaleh

Abstract:

A crucial component of maintaining a customer-oriented business as in the telecom industry is understanding the reasons and factors that lead to customer churn. Competition between telecom companies has greatly increased in recent years. It has become more important to understand customers’ needs in this strong market of telecom industries, especially for those who are looking to turn over their service providers. So, predictive churn is now a mandatory requirement for retaining those customers. Machine learning can be utilized to accomplish this. Churn Prediction has become a very important topic in terms of machine learning classification in the telecommunications industry. Understanding the factors of customer churn and how they behave is very important to building an effective churn prediction model. This paper aims to predict churn and identify factors of customers’ churn based on their past service usage history. Aiming at this objective, the study makes use of feature selection, normalization, and feature engineering. Then, this study compared the performance of four different machine learning algorithms on the Orange dataset: Logistic Regression, Random Forest, Decision Tree, and Gradient Boosting. Evaluation of the performance was conducted by using the F1 score and ROC-AUC. Comparing the results of this study with existing models has proven to produce better results. The results showed the Gradients Boosting with feature selection technique outperformed in this study by achieving a 99% F1-score and 99% AUC, and all other experiments achieved good results as well.

Keywords: machine learning, gradient boosting, logistic regression, churn, random forest, decision tree, ROC, AUC, F1-score

Procedia PDF Downloads 134

Authors: Bliss Singhal

Abstract:

Machine learning (ML) is a branch of Artificial Intelligence (AI) where computers analyze data and find patterns in the data. The study focuses on the detection of metastatic cancer using ML. Metastatic cancer is the stage where cancer has spread to other parts of the body and is the cause of approximately 90% of cancer-related deaths. Normally, pathologists spend hours each day to manually classifying whether tumors are benign or malignant. This tedious task contributes to mislabeling metastasis being over 60% of the time and emphasizes the importance of being aware of human error and other inefficiencies. ML is a good candidate to improve the correct identification of metastatic cancer, saving thousands of lives and can also improve the speed and efficiency of the process, thereby taking fewer resources and time. So far, the deep learning methodology of AI has been used in research to detect cancer. This study is a novel approach to determining the potential of using preprocessing algorithms combined with classification algorithms in detecting metastatic cancer. The study used two preprocessing algorithms: principal component analysis (PCA) and the genetic algorithm, to reduce the dimensionality of the dataset and then used three classification algorithms: logistic regression, decision tree classifier, and k-nearest neighbors to detect metastatic cancer in the pathology scans. The highest accuracy of 71.14% was produced by the ML pipeline comprising of PCA, the genetic algorithm, and the k-nearest neighbor algorithm, suggesting that preprocessing and classification algorithms have great potential for detecting metastatic cancer.

Keywords: breast cancer, principal component analysis, genetic algorithm, k-nearest neighbors, decision tree classifier, logistic regression

Procedia PDF Downloads 83

Authors: Christos E. Lampropoulos, Maria Konsta, Tamta Sirbilatze, Ifigenia Apostolou, Vicky Dradaki, Konstantina Panouria, Irini Dri, Christina Kordali, Vaggelis Lambas, Georgios Mavras

Abstract:

Objectives: Full Mini Nutritional Assessment (MNA) questionnaire is one of the most useful tools in diagnosis of malnutrition in hospitalized patients, which is related to increased morbidity and mortality. The purpose of our study was to assess the nutritional status of elderly, hospitalized patients and examine the hypothesis that MNA may predict mortality and extension of hospitalization. Methods: One hundred fifty patients (78 men, 72 women, mean age 80±8.2) were included in this cross-sectional study. The following data were taken into account in analysis: anthropometric and laboratory data, physical activity (International Physical Activity Questionnaires, IPAQ), smoking status, dietary habits, cause and duration of current admission, medical history (co-morbidities, previous admissions). Primary endpoints were mortality (from admission until 6 months afterwards) and duration of admission. The latter was compared to national guidelines for closed consolidated medical expenses. Logistic regression and linear regression analysis were performed in order to identify independent predictors for mortality and extended hospitalization respectively. Results: According to MNA, nutrition was normal in 54/150 (36%) of patients, 46/150 (30.7%) of them were at risk of malnutrition and the rest 50/150 (33.3%) were malnourished. After performing multivariate logistic regression analysis we found that the odds of death decreased 20% per each unit increase of full MNA score (OR=0.8, 95% CI 0.74-0.89, p < 0.0001). Patients who admitted due to cancer were 23 times more likely to die, compared to those with infection (OR=23, 95% CI 3.8-141.6, p=0.001). Similarly, patients who admitted due to stroke were 7 times more likely to die (OR=7, 95% CI 1.4-34.5, p=0.02), while these with all other causes of admission were less likely (OR=0.2, 95% CI 0.06-0.8, p=0.03), compared to patients with infection. According to multivariate linear regression analysis, each increase of unit of full MNA, decreased the admission duration on average 0.3 days (b:-0.3, 95% CI -0.45 - -0.15, p < 0.0001). Patients admitted due to cancer had on average 6.8 days higher extension of hospitalization, compared to those admitted for infection (b:6.8, 95% CI 3.2-10.3, p < 0.0001). Conclusion: Mortality and extension of hospitalization is significantly increased in elderly, malnourished patients. Full MNA score is a useful diagnostic tool of malnutrition.

Keywords: duration of admission, malnutrition, mini nutritional assessment score, prognostic factors for mortality

Procedia PDF Downloads 313

Authors: Yi-Cheng Tian, Miin-Shen Yang

Abstract:

The mixture likelihood approach to clustering is a popular clustering method where the expectation and maximization (EM) algorithm is the most used mixture likelihood method. In the literature, the EM algorithm had been used for mixture regression models. However, these EM mixture regression algorithms are sensitive to initial values with a priori number of clusters. In this paper, to resolve these drawbacks, we construct a learning-based schema for the EM mixture regression algorithm such that it is free of initializations and can automatically obtain an approximately optimal number of clusters. Some numerical examples and comparisons demonstrate the superiority and usefulness of the proposed learning-based EM mixture regression algorithm.

Keywords: clustering, EM algorithm, Gaussian mixture model, mixture regression model

Procedia PDF Downloads 510

Authors: Lars Laußat, Manfred Helmus, Kamil Szczesny, Markus König

Abstract:

As one result of the project “Reactive Construction Project Scheduling using Real Time Construction Logistic Data and Simulation”, a procedure for using data about uncertain resource availability assumptions in reactive scheduling processes has been developed. Prediction data about resource availability is generated in a formalized way using real-time monitoring data e.g. from auto-ID systems on the construction site and in the supply chains. The paper focuses on the formalization of the procedure for monitoring construction logistic processes, for the detection of disturbance and for generating of new and uncertain scheduling assumptions for the reactive resource constrained simulation procedure that is and will be further described in other papers.

Keywords: auto-ID, construction logistic, fuzzy, monitoring, RFID, scheduling

Procedia PDF Downloads 516

Authors: Anjelika Kremer, Irina Nachimov, Dan Justo

Abstract:

Background: C-reactive protein (CRP) serum levels are commonly measured in hospitalized patients. Elevated admission CRP serum levels and in-hospital mortality has been seldom studied in the general population of elderly patients admitted to the acute Geriatrics department. Methods: A retrospective cross-sectional study was conducted at a tertiary medical center. Included were all elderly patients (age 65 years or more) admitted to a single acute Geriatrics department from the emergency room between April 2014 and January 2015. CRP serum levels were measured routinely in all patients upon the first 24 hours of admission. A logistic regression analysis was used to study if admission CRP serum levels were associated with in-hospital mortality independent of age, gender, functional status, and co-morbidities. Results: Overall, 498 elderly patients were included in the analysis: 306 (61.4%) female patients and 192 (38.6%) male patients. The mean age was 84.8±7.0 years (median: 85 years; IQR: 80-90 years). The mean admission CRP serum levels was 43.2±67.1 mg/l (median: 13.1 mg/l; IQR: 2.8-51.7 mg/l). Overall, 33 (6.6%) elderly patients died during the hospitalization. A logistic regression analysis showed that in-hospital mortality was independently associated with history of stroke (p < 0.0001), heart failure (p < 0.0001), and admission CRP serum levels (p < 0.0001) – and to a lesser extent with age (p = 0.042), collagen vascular disease (p=0.011), and recent venous thromboembolism (p=0.037). Receiver operating characteristic (ROC) curve showed that admission CRP serum levels predict in-hospital mortality fairly with an area under the curve (AUC) of 0.694 (p < 0.0001). Cut-off value with maximal sensitivity and specificity was 19.7 mg/L. Conclusions: Admission CRP serum levels may be used to predict in-hospital mortality in the general population of elderly patients admitted to the acute Geriatrics department.

Keywords: c-reactive protein, elderly, mortality, prediction

Procedia PDF Downloads 239

Authors: Jiaqian Yin, Ruoling Chen, David Churchill, Huijuan Zou, Peipei Guo, Chunmei Liang, Xiaoqing Peng, Zhikang Zhang, Weiju Zhou, Yunxia Cao

Abstract:

Purpose: This study aimed to explore the interaction of male and female age on implantation failure from in vitro fertilisation (IVF)/ intracytoplasmic sperm injection (ICSI) treatments in couples following their first cycles using the Anhui Maternal-Child Health Study (AMCHS). Methods: The AMCHS recruited 2042 infertile couples who were physically fit for in vitro fertilisation (IVF) or intracytoplasmic sperm injection (ICSI) treatment at the Reproductive Centre of the First Affiliated Hospital of Anhui Medical University between May 2017 to April 2021. This prospective cohort study analysed the data from 1910 cohort couples for the current paper data analysis. The multivariate logistic regression model was used to identify the effect of male and female age on implantation failure after controlling for confounding factors. Male age and female age were examined as continuous and categorical (male age: 20-<25, 25-<30, 30-<35, 35-<40, ≥40; female age: 20-<25, 25-<30, 30-<35, 35-<40, ≥40) predictors. Results: Logistic regression indicated that advanced maternal age was associated with increased implantation failure (P<0.001). There was evidence of an interaction between maternal age (30-<35 and ≥ 35) and paternal age (≥35) on implantation failure. (p<0.05). Only when the male was ≥35 years of increased maternal age was associated with the risk of implantation failure. Conclusion: In conclusion, there was an additive effect on implantation failure with advanced parental age. The impact of advanced maternal age was only seen in the older paternal age group. The delay of childbearing in both men and women will be a serious public issue that may contribute to a higher risk of implantation failure in patients needing assisted reproductive technology (ART).

Keywords: parental age, infertility, cohort study, IVF

Procedia PDF Downloads 155

Authors: Sonal Sethi, Mukesh Yadav, Abhimanyu Gupta

Abstract:

The upcoming field of radiogenomics has the potential to upgrade the role of imaging in lung cancer management by noninvasive characterization of tumor histology and genetic microenvironment. Receptor positivity like epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) genotyping are critical in lung adenocarcinoma for treatment. As conventional identification of receptor positivity is an invasive procedure, we analyzed the features on non-invasive computed tomography (CT), which predicts the receptor positivity in lung adenocarcinoma. Retrospectively, we did a comprehensive study from 77 proven lung adenocarcinoma patients with CT images, EGFR and ALK receptor genotyping, and clinical information. Total 22/77 patients were receptor-positive (15 had only EGFR mutation, 6 had ALK mutation, and 1 had both EGFR and ALK mutation). Various morphological characteristics and metastatic distribution on CT were analyzed along with the clinical information. Univariate and multivariable logistic regression analyses were used. On multivariable logistic regression analysis, we found spiculated margin, lymphangitic spread, air bronchogram, pleural effusion, and distant metastasis had a significant predictive value for receptor mutation status. On univariate analysis, air bronchogram and pleural effusion had significant individual predictive value. Conclusions: Receptor positive lung cancer has characteristic imaging features compared with nonreceptor positive lung adenocarcinoma. Since CT is routinely used in lung cancer diagnosis, we can predict the receptor positivity by a noninvasive technique and would follow a more aggressive algorithm for evaluation of distant metastases as well as for the treatment.

Keywords: lung cancer, multidisciplinary cancer care, oncologic imaging, radiobiology

Procedia PDF Downloads 136

Authors: Zerubabel Mihret

Abstract:

Background: Work-related stress is a pattern of reactions to work demands unmatched by worker’s knowledge, skills, or abilities. Healthcare institutions are considered high-risk and intensive work areas for work-related stress. However, there is the nonexistence of clear and strong data about the magnitude of work-related stress on sanitation workers in hospitals in Ethiopia. The aim of this study was to determine the magnitude of work-related stress among sanitation workers in public hospitals during COVID-19 in Addis Ababa, Ethiopia. Methods: Institution-based cross-sectional study was conducted from October 2021 to February 2022 among 494 sanitation workers who were selected from 4 hospitals. HSE (Health and Safety Executive of UK) standard data collection tool was used, and an interviewer-administered questionnaire was used to collect the data using KOBO collect application. The collected data were cleaned and analyzed using SPSS version 20.0. Both binary and multivariable logistic regression analyses were done to identify important factors having an association with work-related stress. Variables with p-value ≤ 0.25 in the bivariate analysis were entered into the multivariable logistic regression model. A statistically significant level was declared at a p-value ≤ 0.05. Results: This study revealed that the magnitude of work-related stress among sanitation workers was 49.2% (95% CI 45-54). Significant proportions (72.7%) of sanitation workers were dissatisfied with their current job. Sex, age, experience, and chewing khat were significantly associated with work-related stress. Conclusion: Work-related stress is significantly high among sanitation workers. Sex, age, experience, and chewing khat were identified as factors associated with work-related stress. Intervention program focusing on the prevention and control of stress is desired by hospitals.

Keywords: work-related stress, sanitation workers, Likert scale, public hospitals, Ethiopia

Procedia PDF Downloads 86

Authors: Lea Stein, Sabine Rösner, Alessandra Lo Giudice, Beate Ditzen, Tewes Wischmann

Abstract:

For many couples bearing children is a consistent life goal; however, it cannot always be fulfilled. Undergoing infertility treatment does not guarantee pregnancies and live births. Couples have to deal with miscarriages and sometimes even discontinue infertility treatment. Significant medical predictors for the outcome of infertility treatment have yet to be fully identified. To further our understanding, a cross-sectional five-year follow-up survey was undertaken, in which 95 women and 82 men that have been treated at the Women’s Hospital of Heidelberg University participated. Binary logistic regressions, parametric and non-parametric methods were used for our sample to determine the relevance of biological (infertility diagnoses, maternal and paternal age) and lifestyle factors (smoking, drinking, over- and underweight) on the outcome of infertility treatment (clinical pregnancy, live birth, miscarriage, dropout rate). During infertility treatment, 72.6% of couples became pregnant and 69.5% were able to give birth. Suffering from miscarriages 27.5% of couples and 20.5% decided to discontinue an unsuccessful fertility treatment. The binary logistic regression models for clinical pregnancies, live births and dropouts were statistically significant for the maternal age, whereas the paternal age in addition to maternal and paternal BMI, smoking, infertility diagnoses and infections, showed no significant predicting effect on any of the outcome variables. The results confirm an effect of maternal age on infertility treatment, whereas the relevance of other medical predictors remains unclear. Further investigations should be considered to increase our knowledge of medical predictors.

Keywords: advanced maternal age, assisted reproductive technology, female factor, male factor, medical predictors, infertility treatment, reproductive medicine

Procedia PDF Downloads 111

Authors: Sheena K. J., Jyoti Badge, Sayed Mohammed Zeeshan

Abstract:

A Comparative Study of Exponential and Logistic Population Growth Models in India India is the second most populous city in the world, just behind China, and is going to be in the first place by next year. The Indian population has remarkably at higher rate than the other countries from the past 20 years. There were many scientists and demographers who has formulated various models of population growth in order to study and predict the future population. Some of the models are Fibonacci population growth model, Exponential growth model, Logistic growth model, Lotka-Volterra model, etc. These models have been effective in the past to an extent in predicting the population. However, it is essential to have a detailed comparative study between the population models to come out with a more accurate one. Having said that, this research study helps to analyze and compare the two population models under consideration - exponential and logistic growth models, thereby identifying the most effective one. Using the census data of 2011, the approximate population for 2016 to 2031 are calculated for 20 Indian states using both the models, compared and recorded the data with the actual population. On comparing the results of both models, it is found that logistic population model is more accurate than the exponential model, and using this model, we can predict the future population in a more effective way. This will give an insight to the researchers about the effective models of population and how effective these population models are in predicting the future population.

Keywords: population growth, population models, exponential model, logistic model, fibonacci model, lotka-volterra model, future population prediction, demographers

Procedia PDF Downloads 124

Authors: Brindha Senthil Kumar, Payel Chakraborty, Senthil Kumar Nachimuthu, Arindam Maitra, Prem Nath

Abstract:

Gastric cancer is predominantly caused by demographic and diet factors as compared to other cancer types. The aim of the study is to predict Early Gastric Cancer (ECG) from diet and lifestyle factors using supervised machine learning algorithms. For this study, 160 healthy individual and 80 cases were selected who had been followed for 3 years (2016-2019), at Civil Hospital, Aizawl, Mizoram. A dataset containing 11 features that are core risk factors for the gastric cancer were extracted. Supervised machine algorithms: Logistic Regression, Naive Bayes, Support Vector Machine (SVM), Multilayer perceptron, and Random Forest were used to analyze the dataset using Python Jupyter Notebook Version 3. The obtained classified results had been evaluated using metrics parameters: minimum_false_positives, brier_score, accuracy, precision, recall, F1_score, and Receiver Operating Characteristics (ROC) curve. Data analysis results showed Naive Bayes - 88, 0.11; Random Forest - 83, 0.16; SVM - 77, 0.22; Logistic Regression - 75, 0.25 and Multilayer perceptron - 72, 0.27 with respect to accuracy and brier_score in percent. Naive Bayes algorithm out performs with very low false positive rates as well as brier_score and good accuracy. Naive Bayes algorithm classification results in predicting ECG showed very satisfactory results using only diet cum lifestyle factors which will be very helpful for the physicians to educate the patients and public, thereby mortality of gastric cancer can be reduced/avoided with this knowledge mining work.

Keywords: Early Gastric cancer, Machine Learning, Diet, Lifestyle Characteristics

Procedia PDF Downloads 163

Authors: Nur Atiqah Adam, Mor Jack Ng, Bernard Chern, Kok Hian Tan

Abstract:

Introduction: Gestational diabetes (GDM) is a common pregnancy complication with short and long-term health consequences for both mother and fetus. Factors such as family history of diabetes mellitus, maternal obesity, maternal age, ethnicity and parity have been reported to influence the risk of GDM. In a multi-racial country like Singapore, it is worthwhile to study the GDM prevalences of different ethnicities. We aim to investigate the influence of ethnicity on the racial prevalences of GDM in Singapore. This is important as it may help us to improve guidelines on GDM healthcare services according to significant risk factors unique to Singapore. Materials and Methods: Obstetric cohort data of 926 singleton deliveries in KK Women’s and Children’s Hospital (KKH) from 2011 to 2013 was obtained. Only patients aged 18 and above and without complicated pregnancies or chronic illnesses were targeted. Factors such as ethnicity, maternal age, parity and maternal body mass index (BMI) at booking visit were studied. A multivariable logistic regression model, adjusted for confounders, was used to determine which of these factors are significantly associated with an increased risk of GDM. Results: The overall GDM prevalence rate based on WHO 1999 criteria & at risk screening (race alone not a risk factor) was 8.86%. GDM rates were higher among women above 35 years old (15.96%), obese (15.15%) and multiparous women (10.12%). Indians had a higher GDM rate (13.0 %) compared to the Chinese (9.57%) and Malays (5.20%). However, using multiple logistic regression model, variables that are significantly related to GDM rates were maternal age (p < 0.001) and maternal BMI at booking visit (p = 0.006). Conclusion: Maternal age (p < 0.001) and maternal booking BMI (p = 0.006) are the strongest risk factors for GDM. Ethnicity per se does not seem to have a significant influence on the prevalence of GDM in Singapore (p = 0.064). Hence we should tailor guidelines on GDM healthcare services according to maternal age and booking BMI rather than ethnicity.

Keywords: ethnicity, gestational diabetes, healthcare, pregnancy

Procedia PDF Downloads 226

Authors: Kisan Sarda, Bhavika Shingote

Abstract:

This paper aims to evaluate regression modelling for prediction of Energy storage of solar photovoltaic (PV) system using Semi parametric regression techniques because there are some parameters which are known while there are some unknown parameters like humidity, dust etc. Here irradiation of solar energy is different for different places on the basis of Latitudes, so by finding out areas which give more storage we can implement PV systems at those places and our need of energy will be fulfilled. This regression modelling is done for daily, monthly and seasonal prediction of solar energy storage. In this, we have used R modules for designing the algorithm. This algorithm will give the best comparative results than other regression models for the solar PV cell energy storage.

Keywords: semi parametric regression, photovoltaic (PV) system, regression modelling, irradiation

Procedia PDF Downloads 382

Authors: Yahaya Abdullahi, Ben Coetzee, Linda Van Den Berg

Abstract:

The study aimed at establishing the global positioning system (GPS) determined singles match characteristics that act as predictors of successful and less-successful male singles badminton players’ group classification. Twenty-two (22) male single players (aged: 23.39 ± 3.92 years; body stature: 177.11 ± 3.06cm; body mass: 83.46 ± 14.59kg) who represented 10 African countries participated in the study. Players were categorised as successful and less-successful players according to the results of five championships’ of the 2014/2015 season. GPS units (MinimaxX V4.0), Polar Heart Rate Transmitter Belts and digital video cameras were used to collect match data. GPS-related variables were corrected for match duration and independent t-tests, a cluster analysis and a binary forward stepwise logistic regression were calculated. A Receiver Operating Characteristic Curve (ROC) was used to determine the validity of the group classification model. High-intensity accelerations per second were identified as the only GPS-determined variable that showed a significant difference between groups. Furthermore, only high-intensity accelerations per second (p=0.03) and low-intensity efforts per second (p=0.04) were identified as significant predictors of group classification with 76.88% of players that could be classified back into their original groups by making use of the GPS-based logistic regression formula. The ROC showed a value of 0.87. The identification of the last-mentioned GPS-related variables for the attainment of badminton performances, emphasizes the importance of using badminton drills and conditioning techniques to not only improve players’ physical fitness levels but also their abilities to accelerate at high intensities.

Keywords: badminton, global positioning system, match analysis, inertial movement analysis, intensity, effort

Procedia PDF Downloads 192

Authors: Zhuoran Shan, Li Wan, Xianchun Zhang

Abstract:

With an increasing trend of regionalization and polycentricity in Chinese contemporary big cities, “sub-region life circle” turns to be an effective method on rational organization of urban function and spatial structure. By the method of questionnaire, network big data, route inversion on internet map, GIS spatial analysis and logistic regression, this article makes research on characteristic and formation mechanism of “sub-region life circle” based on human high-probability daily activity in Chinese big cities. Firstly, it shows that “sub-region life circle” has been a new general spatial sphere of residents' high-probability daily activity and mobility in China. Unlike the former analysis of the whole metropolitan or the micro community, “sub-region life circle” has its own characteristic on geographical sphere, functional element, spatial morphology and land distribution. Secondly, according to the analysis result with Binary Logistic Regression Model, the research also shows that seven factors including land-use mixed degree and bus station density impact the formation of “sub-region life circle” most, and then analyzes the index critical value of each factor. Finally, to establish a smarter “sub-region life circle”, this paper indicates that several strategies including jobs-housing fit, service cohesion and space reconstruction are the keys for its spatial organization optimization. This study expands the further understanding of cities' inner sub-region spatial structure based on human daily activity, and contributes to the theory of “life circle” in urban's meso-scale.

Keywords: sub-region life circle, characteristic, formation mechanism, human activity, spatial structure

Procedia PDF Downloads 301

Authors: Fernanda M. Assef, Maria Teresinha A. Steiner

Abstract:

The aim of this paper is to compare and discuss better classifier algorithm options for credit risk assessment by applying different Machine Learning techniques. Using records from a Brazilian financial institution, this study uses a database of 5,432 companies that are clients of the bank, where 2,600 clients are classified as non-defaulters, 1,551 are classified as defaulters and 1,281 are temporarily defaulters, meaning that the clients are overdue on their payments for up 180 days. For each case, a total of 15 attributes was considered for a one-against-all assessment using four different techniques: Artificial Neural Networks Multilayer Perceptron (ANN-MLP), Artificial Neural Networks Radial Basis Functions (ANN-RBF), Logistic Regression (LR) and finally Support Vector Machines (SVM). For each method, different parameters were analyzed in order to obtain different results when the best of each technique was compared. Initially the data were coded in thermometer code (numerical attributes) or dummy coding (for nominal attributes). The methods were then evaluated for each parameter and the best result of each technique was compared in terms of accuracy, false positives, false negatives, true positives and true negatives. This comparison showed that the best method, in terms of accuracy, was ANN-RBF (79.20% for non-defaulter classification, 97.74% for defaulters and 75.37% for the temporarily defaulter classification). However, the best accuracy does not always represent the best technique. For instance, on the classification of temporarily defaulters, this technique, in terms of false positives, was surpassed by SVM, which had the lowest rate (0.07%) of false positive classifications. All these intrinsic details are discussed considering the results found, and an overview of what was presented is shown in the conclusion of this study.

Keywords: artificial neural networks (ANNs), classifier algorithms, credit risk assessment, logistic regression, machine Learning, support vector machines

Procedia PDF Downloads 104

Authors: Yi-Hsien Wang, Fu-Ju Yang, Hwa-Rong Shen, Rui-Lin Tseng

Abstract:

The business is notably related to the market risk according to the increase of liberalization of financial markets. Hence, the company usually utilized high financial leverage of derivatives to hedge the risk. When the company choose different hedging instruments to face a variety of exchange rate risk, we employ the Multinomial Logistic-AHP to analyze the impact of various derivatives. Hence, the research summarized the literature on relevant factors affecting managers selected exchange rate hedging instruments, using Multinomial Logistic Model and and further integrate AHP. Using Experts’ Questionnaires can test multi-level selection and hedging effect of different hedging instruments in order to calculate the hedging instruments and the multi-level factors of weights to understand the gap between the empirical results and practical operation. Finally, the Multinomial Logistic-AHP Model will sort the weights to analyze. The research findings can be a basis reference for investors in decision-making.

Keywords: exchange rate risk, derivatives, hedge, multinomial logistic-AHP

Procedia PDF Downloads 443

Authors: Mukesh Ravi Raushan

Abstract:

Despite consideration of obesity as a deadly public health issue contributing 2.6 million deaths worldwide every year developing country like India is facing malnutrition and it is more common than in Sub-Saharan Africa. About one in every three malnourished children in the world lives in India. The paper assess the nutritional health among children using data from total number of 43737 infant and young children aged 0-59 months (µ = 29.54; SD = 17.21) of the selected households by National Family Health Survey, 2005-06. The wasting was measured by a Z-score of standardized weight-for-height according to the WHO child growth standards. The impact of education with place of residence was found to be significantly associated with the complementary food diversity score (CFDS) in India. The education of mother was positively associated with the CFDS but the degree of performance was lower in rural India than their counterpart from urban. The result of binary logistic regression on wasting with WHO seven types of recommended food for children in India suggest that child who consumed the milk product food (OR: 0.87, p<0.0001) were less likely to be malnourished than their counterparts who did not consume, whereas, in case of other food items as the child who consumed food product of seed (OR: 0.75, p<0.0001) were less likely to be malnourished than those who did not. The nutritional status among children were negatively associated with the protein containing complementary food given the child as those child who received pulse in last 24 hour were less likely to be wasted (OR: 0.87, p<0.00001) as compared to the reference categories. The frequency to feed the indexed child increases by 10 per cent the expected change in child health outcome in terms of wasting decreases by 2 per cent in India when place of residence, education, religion, and birth order were controlled. The index gets improved as the risk for malnutrition among children in India decreases.

Keywords: CFDS, food diversity index, India, logistic regression

Procedia PDF Downloads 262

Authors: Lubena Achmad, Herman Kristanto, Julian Dewantiningrum

Abstract:

Background: The Bishop score is a standard method used to predict the success of induction. This examination tends to be subjective with high inter and intraobserver variability, so it was presumed to have a low predictive value in terms of the outcome of labor induction. Cervical length measurement using transvaginal ultrasound is considered to be more objective to assess the cervical length. Meanwhile, this examination is not a complicated procedure and less invasive than vaginal touché. Objective: To compare transvaginal ultrasound and Bishop score in predicting successful induction. Methods: This study was a prospective cohort study. One hundred and twenty women with singleton pregnancies undergoing induction of labor at 37 – 42 weeks and met inclusion and exclusion criteria were enrolled in this study. Cervical assessment by both transvaginal ultrasound and Bishop score were conducted prior induction. The success of labor induction was defined as an ability to achieve active phase ≤ 12 hours after induction. To figure out the best cut-off point of cervical length and Bishop score, receiver operating characteristic (ROC) curves were plotted. Logistic regression analysis was used to determine which factors best-predicted induction success. Results: This study showed significant differences in terms of age, premature rupture of the membrane, the Bishop score, cervical length and funneling as significant predictors of successful induction. Using ROC curves found that the best cut-off point for prediction of successful induction was 25.45 mm for cervical length and 3 for Bishop score. Logistic regression was performed and showed only premature rupture of membranes and cervical length ≤ 25.45 that significantly predicted the success of labor induction. By excluding premature rupture of the membrane as the indication of induction, cervical length less than 25.3 mm was a better predictor of successful induction. Conclusion: Compared to Bishop score, cervical length using transvaginal ultrasound was a better predictor of successful induction.

Keywords: Bishop Score, cervical length, induction, successful induction, transvaginal sonography

Procedia PDF Downloads 325