Abstract:
Maternal health necessitates proactive measures for risk assessment and management due to its potentially severe consequences for both mothers and babies. Consequently, there is an urgent need to develop accurate and reliable methods to predict maternal health risk levels utilizing easily accessible physiological data. Presently, healthcare professionals depend on traditional risk assessment methods that rely on limited information and subjective evaluations, which may inadequately capture the intricate relationship of multiple factors impacting maternal health. Moreover, these methods are susceptible to human errors and variability in judgment, resulting in suboptimal decision-making and delayed interventions. This research employed a dataset comprising maternal health risk attributes, including age, systolic blood pressure, diastolic blood pressure, blood sugar, body temperature, and heart rate, to develop predictive models for maternal health risk classification. Several machine learning algorithms are applied to the dataset, including Random Forest, Support Vector Machines (SVM), Gradient Boosting, AdaBoost, K-Nearest Neighbors (KNN), Gaussian Naive Bayes, and Light GBM. The dataset was carefully preprocessed using appropriate techniques. The performance of each model was evaluated using various metrics such as accuracy, precision, recall, and F1-score. The models are compared based on their predictive performance, interpretability, and computational efficiency to determine the most effective algorithm for predicting maternal health risk levels. After implementing various machine learning algorithms on the dataset, Light GBM emerged as the top-performing algorithm, exhibiting superior predictive performance for maternal health risk classification with an accuracy of 84.24%. This exceptional accuracy demonstrates the effectiveness of Light GBM in accurately predicting maternal health risk levels. Its ability to handle complex relationships within the data and efficiently process large datasets contributes to its superior performance. The achieved accuracy highlights the potential of machine learning techniques, particularly Light GBM, in supporting healthcare professionals in assessing maternal health risks. Accurate identification of high-risk pregnancies enables the implementation of appropriate interventions and personalized care plans, ultimately enhancing maternal healthcare outcomes. The findings underscore the importance of advanced data analysis techniques in maternal healthcare, with the predictive models offering valuable insights to support healthcare professionals in improving maternal health outcomes. Further validation, refinement, and expansion of the models are recommended to enhance generalizability and real-world applicability. Overall, this research contributes to the advancement of maternal healthcare by demonstrating the effectiveness of Light GBM and its potential to inform decision-making for the well-being of expectant mothers and their babies. Future research should also consider incorporating additional relevant variables and expanding the dataset to strengthen the predictive capabilities of the models.
Page(s):
13-13
DOI:
DOI not available
Published:
Journal: Abstract Book on International Conference on Food and Applied Sciences (ICFAS-23) 3-5 August 23, Volume: 0, Issue: 0, Year: 2023
Keywords:
Maternal Health
,
Maternal health
,
Maternal health
,
risk prediction
,
Multivariate
,
Maternal Health
,
Machine learning algorithms
,
Light GBM