A nested stacking ensemble model for predicting districts with high and low maternal mortality ratio (MMR) in India
| Autor: | Kuljeet Singh, Anand Sharma, Vibhakar Mansotra, Paramjit Kour, Sourabh Shastri, Sachin Kumar |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Computer Networks and Communications
Computer science media_common.quotation_subject Stacking 02 engineering and technology Machine learning computer.software_genre Cross-validation Field (computer science) Artificial Intelligence Voting 0202 electrical engineering electronic engineering information engineering Electrical and Electronic Engineering media_common Ensemble forecasting business.industry Applied Mathematics 020206 networking & telecommunications Computer Science Applications Statistical classification Standardized mortality ratio Computational Theory and Mathematics Healthcare industry 020201 artificial intelligence & image processing Artificial intelligence business computer Information Systems |
| Zdroj: | International Journal of Information Technology. 13:433-446 |
| ISSN: | 2511-2112 2511-2104 |
| DOI: | 10.1007/s41870-020-00560-3 |
| Popis: | The ensemble is an efficacious machine learning framework that combines variety of algorithms for better performance and effective prediction. Over the past few years, numerous researchers proposed wide variety of ensemble methodologies in the field of healthcare industry. In the present research paper, a nested ensemble has been suggested based on Stacking and Voting schemes for prediction and analysis of Maternal Mortality Ratio (MMR) in India. The presented nested ensemble combines Base Learners and Meta Learners by employing different classification algorithms and prediction results were afterwards evaluated by using K-fold cross validation and thus, facilitating the statistical distribution of results. Further, the effectiveness of the ensemble was investigated by comparing its performance with the various single learning algorithms in terms of accuracy, precision, recall, F-measure and ROC. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink