LSTM With Adam Optimization-Powered High Accuracy Preeclampsia Classification
| Autor: | Tessy Badriyah, Muhlis Tahir, Nur Sakinah, Iwan Syarif |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
030506 rehabilitation
Artificial neural network Computer science business.industry Pattern recognition 030229 sport sciences medicine.disease Preeclampsia 03 medical and health sciences 0302 clinical medicine Blood pressure medicine Artificial intelligence 0305 other medical science business reproductive and urinary physiology Organ system Test data |
| Zdroj: | 2019 International Electronics Symposium (IES). |
| DOI: | 10.1109/elecsym.2019.8901536 |
| Popis: | Preeclampsia is a pregnancy complication characterized by high blood pressure and signs of damage to another organ system, most often the liver and kidneys. Preeclampsia usually begins after 20 weeks of pregnancy in women whose blood pressure had been normal. To anticipate the risk of preeclampsia and prevent the death of pregnant women, a predictive system is needed for preeclampsia. to predict this, the data obtained will be studied to recognize patterns for predicting preeclampsia. Artificial neural network is one method that has the ability to learn patterns from a data. The system built in this study is to use the artificial neural network method by using the architecture of Long Short-Term Memory (LSTM) Neural Networks. But this technique requires the right parameters to get accurate predictions. In this paper analyze several parameters such as number of time series patterns, number of hidden neurons, max epoch, and composition of training and test data on the accuracy of the predictions obtained. The results of the analysis showed that the system built was able to predict Preeclampsia well, with an average accuracy rate of 90.22% for data testing. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|