2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine
| Format: | General Document |
|---|
| _version_ | 1860798329944801280 |
|---|---|
| building | INTELEK Repository |
| collection | Online Access |
| collectionurl | https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 |
| copyright | Copyright©PWB2025 |
| country | Malaysia |
| date | 2024-09-09 04:53 |
| format | General Document |
| id | 17235 |
| institution | UniSZA |
| originalfilename | DETERMINANTS OF SHORT INTERPREGNANCY INTERVAL COMPARISON OF PREDICTIVE ACCURACY OF BINARY MULTIPLE LOGISTIC REGRESSION AND SUPPORT VECTOR MACHINE (PHD_2024).pdf |
| person | Windows User |
| recordtype | oai_dc |
| resourceurl | https://intelek.unisza.edu.my/intelek/pages/view.php?ref=17235 |
| sourcemedia | Server storage Scanned document |
| spelling | 17235 https://intelek.unisza.edu.my/intelek/pages/view.php?ref=17235 https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 General Document Malaysia Library Staff (Top Management) Library Staff (Management) Library Staff (Support) Terengganu Faculty of Medicine English application/pdf 1.7 Windows User Server storage Scanned document UniSZA Private Access UniSZA 209 Copyright©PWB2025 Dissertations, Academic Machine Learning in Healthcare Microsoft® Word LTSC UniSZA Pregnancy — Health aspects Maternal Health Services — Evaluation Logistic Regression Models — Application In Medicine Predictive Analytics — Health Sciences Interpregnancy Interval Birth Spacing Maternal Health Predictive Modelling Logistic Regression Support Vector Machine (SVM) Reproductive Health Risk Factors Pregnancy Outcomes 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine INTRODCUTION: Short Inter Pregnancy Interval (SIPI) is a prevalent and critical public health issue especially in Pakistan. Women with SIPI do not get enough time to heal,recover and replenish lost nutrients to sustain another healthy pregnancy which can result in worse feto-maternal outcomes. The literature is limited regarding determinants of SIPI both internationally as well as locally. Moreover, no study has so far compared the predictive accuracy of conventional Logistic Regression (LR) with relatively innovative machine learning techniques, such as Support Vector Machine (SVM). The objective of this study was conducted to identify risk factors of SIPI and compare the predictive accuracy of LR vs SVM in identification of significant predictors of SIPI. METHODOLOGY: This retrospective cohort study design was done to collect data from pregnant females visiting Lady Aitcheson Hospital Lahore for their antenatal check-ups. The data was calculated from 528 females (264 SIPI, 264 NIPI). Before main data collection, two questionnaires were developed and validated for knowledge assessment about birth spacing and family planning. Both EFA and CFA was applied for validation,whereas, ChronBach’s Alpha was used for internal reliability. After validation, these questionnaires, alongwith information about other socio-demographic, clinical, history and family related risk factors. Data was entered and analyzed in SPSS. Univariate Logistic Regression was applied to identify significant variables for Multiple Logistic Regression (MLR), which was applied after meeting all assumptions. MLR was then compared with SVM for their respective predictive ability.RESULTS: The study found that Poor knowledge of birth spacing and family planning, Desire of more male babies, Own desire of more babies, Breastfeeding of less than 1 year, Desire of equal male and female babies, Knowledge of SIPI, Family doesn’t like SIPI, Knowledge of family planning, Parity<3, rural area of residence, current marital age and History of miscarriage were significant predictors of SIPI (p-values<0.05 for all). The overall accuracy for MLR was 83.14 while Sensitivity, Specificity, PPV and NPV was 81.6%, 85.23%, 84.58% and 81.82% respectively with an area under ROC curve as 92.0089%. For SVM the overall accuracy was 94.70%, with Sensitivity, Specificity, PPV and NPV as 95.08%, 94.32%, 94.36% and 95.04% respectively with area under ROC curve as 98.83%. The Precision and recall for MLR was also less than SVM i.e. 84.58% and 81.06% vs. 94.36% and 95.08% respectively. CONCLUSION: Hence, this study concludes that number of modifiable risk factors including poor knowledge of family planning and birth spacing are significantly associated with Short Inter Pregnancy Interval. Moreover, we conclude that SVM is a fastidious, comprehensive, interactive, flexible and accurate machine learning tool that can be used for better predictions of risk factors of SIPI compared to LR. 2024-09-09 04:53 uuid:01CCD553-3A3C-43F3-9D65-D95EDA77DA74 DETERMINANTS OF SHORT INTERPREGNANCY INTERVAL COMPARISON OF PREDICTIVE ACCURACY OF BINARY MULTIPLE LOGISTIC REGRESSION AND SUPPORT VECTOR MACHINE (PHD_2024).pdf Thesis |
| spellingShingle | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| state | Terengganu |
| subject | Dissertations, Academic Pregnancy — Health aspects Maternal Health Services — Evaluation Logistic Regression Models — Application In Medicine Predictive Analytics — Health Sciences |
| summary | INTRODCUTION: Short Inter Pregnancy Interval (SIPI) is a prevalent and critical public health issue especially in Pakistan. Women with SIPI do not get enough time to heal,recover and replenish lost nutrients to sustain another healthy pregnancy which can result in worse feto-maternal outcomes. The literature is limited regarding determinants of SIPI both internationally as well as locally. Moreover, no study has so far compared the predictive accuracy of conventional Logistic Regression (LR) with relatively innovative machine learning techniques, such as Support Vector Machine (SVM). The objective of this study was conducted to identify risk factors of SIPI and compare the predictive accuracy of LR vs SVM in identification of significant predictors of SIPI. METHODOLOGY: This retrospective cohort study design was done to collect data from pregnant females visiting Lady Aitcheson Hospital Lahore for their antenatal check-ups. The data was calculated from 528 females (264 SIPI, 264 NIPI). Before main data collection, two questionnaires were developed and validated for knowledge assessment about birth spacing and family planning. Both EFA and CFA was applied for validation,whereas, ChronBach’s Alpha was used for internal reliability. After validation, these questionnaires, alongwith information about other socio-demographic, clinical, history and family related risk factors. Data was entered and analyzed in SPSS. Univariate Logistic Regression was applied to identify significant variables for Multiple Logistic Regression (MLR), which was applied after meeting all assumptions. MLR was then compared with SVM for their respective predictive ability.RESULTS: The study found that Poor knowledge of birth spacing and family planning, Desire of more male babies, Own desire of more babies, Breastfeeding of less than 1 year, Desire of equal male and female babies, Knowledge of SIPI, Family doesn’t like SIPI, Knowledge of family planning, Parity<3, rural area of residence, current marital age and History of miscarriage were significant predictors of SIPI (p-values<0.05 for all). The overall accuracy for MLR was 83.14 while Sensitivity, Specificity, PPV and NPV was 81.6%, 85.23%, 84.58% and 81.82% respectively with an area under ROC curve as 92.0089%. For SVM the overall accuracy was 94.70%, with Sensitivity, Specificity, PPV and NPV as 95.08%, 94.32%, 94.36% and 95.04% respectively with area under ROC curve as 98.83%. The Precision and recall for MLR was also less than SVM i.e. 84.58% and 81.06% vs. 94.36% and 95.08% respectively. CONCLUSION: Hence, this study concludes that number of modifiable risk factors including poor knowledge of family planning and birth spacing are significantly associated with Short Inter Pregnancy Interval. Moreover, we conclude that SVM is a fastidious, comprehensive, interactive, flexible and accurate machine learning tool that can be used for better predictions of risk factors of SIPI compared to LR. |
| title | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| title_full | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| title_fullStr | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| title_full_unstemmed | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| title_short | 2024_Determinants Of Short Interpregnancy Interval: Comparison Of Predictive Accuracy Of Multiple Logistic Regression And Support Vector Machine |
| title_sort | 2024_determinants of short interpregnancy interval: comparison of predictive accuracy of multiple logistic regression and support vector machine |