|
Evaluation of Machine Learning Algorithms on the Prediction of Live Birth Occurrence
|
|
| Author:
|
GOWRAMMA G S, DR SHANTHARAM NAYAK, DR NAGARAJ CHOLLI
|
| Abstract:
|
The scientific prediction of the Invitro fertilization (IVF) Live Birth Occurrence process is becoming an
essential medical knowledge, which enables the doctor and the candidate couple to have the prior information
about the health condition and therefore, allows them to take the appropriate step. Prediction of IVF Live
Birth Occurrence is impersonating bigger challenges in obstetrics and gynaecological studies because many
factors such as Intrinsic and Extrinsic factors influence the IVF success rates. In recent past studies machine
learning (ML)model gained huge importance in the IVF success rates prediction. ML techniques have advantage
over other mathematical and conventional method that the ML technique can consider many factors and can
effectively study the interaction among all the significant factors that contribute the success of the IVF Live
birth occurrence. Present study selected three best performing ML models from the literature i.e Gradient
Boosting Regression (GBR), Lasso Regression (LaR) and Linear regression (LR) and compared their accuracy
and efficacy percentage in predicting the IVF live birth occurrence rates. Study constructed DATA from the
literature review on identified significant attributes of both Intrinsic and Extrinsic factors such as age of
women, duration of infertility, egg number which have a direct (or) indirect impact on IVF success rates.
Selected data was used to train all the three different ML algorithms individually to assess and identify the most
reliable ML method employed in the prediction of IVF Live Birth Occurrence rates. The results envisaged that
the LR showed an accuracy of 41%, the LaR showed an accuracy of 40%, and GBR showed an accuracy of 78%.
The present study concludes that among the three different ML algorithms tested GBR algorithm shown the
highest accuracy in both training and testing of IVF Live birth occurrence rates prediction.
|
| Keyword:
|
IVF success rate prediction, Machine learning, Gradient boosting algorithm, Linear regression, Lasso algorithm, Training accuracy, Testing accuracy
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2021.13.02.411
|
| Download:
|
Request For Article
|
|
|