Claim Missing Document
Check
Articles

Found 2 Documents
Search

Optimization of the XG-Boost Algorithm for Predicting Stroke Patient Care Outcomes Puji Lestari; Imam Tahyudin; Ades Tikaningsih; Ade Nurhopipah
Telematika Vol 18, No 1: February (2025)
Publisher : Universitas Amikom Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35671/telematika.v18i1.2817

Abstract

Stroke is a critical health issue in Indonesia, contributing to high mortality rates. At Banyumas District Hospital, stroke is the fourth most common condition, presenting significant challenges in both clinical care and financial management. The purpose of this study is to enhance the quality of services and optimize treatment costs for stroke patients by developing a predictive model using the XGBoost algorithm. This study employs the XGBoost algorithm to develop predictive models, which are then implemented within a web-based machine learning application using the Paython Flask framework. The models predict patient mortality and hospitalization duration. The results indicate that the XGBoost algorithm predicts patient mortality with 86% accuracy and hospitalization duration with 82% accuracy. The developed application significantly enhances care quality and resource management at Banyumas District Hospital by providing accurate predictions to support healthcare decision-making. The use of this application can significantly improve the management of stroke patient care at Banyumas District Hospital, thus maximizing service quality and optimizing treatment costs. By integrating accurate predictive modeling into healthcare decision-making processes, the application facilitates more effective allocation of resources and timely medical interventions, ultimately contributing to better patient outcomes.
Pelatihan Aplikasi Hand Craft Untuk Nasyiatul Aisyiyah Debby Ummul Hidayah; Masyruri Rizka Maulana; Puji Lestari
ABSYARA: Jurnal Pengabdian Pada Masayarakat Vol 3 No 1 (2022): ABSYARA: Jurnal Pengabdian Pada Masyarakat
Publisher : Universitas Hamzanwadi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29408/ab.v3i1.5835

Abstract

Improving technological competence in an organization can be done in various ways. One of them can be done by conducting training on the technology. Like the Rabak Branch of the Nasyiatul 'Aisyiyah organization, related technology training, until now, is still rarely done. One of the activities that technological tools can support is making handicrafts. However, the activity of making handicrafts is still done traditionally. Often do not arrive at the expected result. This condition makes the activity of making handicrafts not optimal. Even the modules used are usually unavailable, so each member of Nasyiatul 'Aisyiyah cannot finish making these handicrafts after the activity. Therefore, this community service activity aims to introduce android-based technology in the form of handicraft applications and to help Nasyiatul 'Aisyiyah members make handicrafts. The implementation methods in this community service activity include giving lectures explaining handcraft applications, assisting participants, and direct practice using handcraft applications. The activity is carried out offline, which is located in Bustanul Athfal 'Aisyiyah Rabak. The steps in this community service activity start from preparation, implementation, and evaluation. The implementation of this community service activity indicates that the handcraft application training activities are included in good criteria, with the evaluation value from the results of the questionnaire calculation showing the figure of 79.58%. This shows that most of the participants have been able to understand and use handcraft applications in making handicrafts