Case Base Reasoning for Diagnosing the Level of Hyperemesis Gravidarum in Pregnant Women using K-Nearest Neighbor
Autor: | Novianti Puspitasari, Ervina Rahayu, Herman Santoso Pakpahan, Medi Taruk, Haviluddin Haviluddin |
---|---|
Jazyk: | indonéština |
Rok vydání: | 2024 |
Předmět: | |
Zdroj: | Sistemasi: Jurnal Sistem Informasi, Vol 13, Iss 3, Pp 874-888 (2024) |
Druh dokumentu: | article |
ISSN: | 2302-8149 2540-9719 |
DOI: | 10.32520/stmsi.v13i3.2584 |
Popis: | Hyperemesis gravidarum is a disease that causes excessive nausea and vomiting in pregnant women. Due to dehydration, this disease can interfere with daily work and get worse. Medical personnel generally recognize hyperemesis gravidarum as one type of disease. In fact, hyperemesis gravidarum is divided into 3 levels, namely grade I or general hyperemesis gravidarum, grade II hyperemesis gravidarum and grade III. This shows that information about hyperemesis gravidarum has yet to be widely known by some medical personnel. If this is left untreated, these two conditions can cause deep vein thrombosis in pregnant women. This study aims to apply the Case-Based Reasoning and K-Nearest Neighbor (KNN) methods to produce accurate information on the diagnosis of hyperemesis gravidarum levels in pregnant women based on symptom management in cases of an old diagnosis. The study used medical record data for hyperemesis gravidarum sufferers in 2018-2019, totalling 228 data. The calculation results of the Case-Based Reasoning method with the K-Nearest Neighbor using the confusion matrix produce an accuracy value of 74%, a precision value of 55% and a recall value of 57%, which indicates that this method is good enough to diagnose levels in patients with hyperemesis gravidarum. |
Databáze: | Directory of Open Access Journals |
Externí odkaz: |