Data Science: Journal of Computing and Applied Informatics

Spatial Clustering Analysis of Stunting in North Sumatra Based on Environmental Factors Using K-Means Algorithm

2024-07-15T10:05:14+07:00

This research aims to analyze the spatial grouping of stunting events in North Sumatra based on environmental factors using the K-Means algorithm. The data used in this research includes the incidence of stunting, environmental factors (such as access to health services, living environment conditions, water use and sanitation), and spatial data (geographical coordinates). The data comes from Basic Health Research (RISKESDAS 2018, then processed and normalized. The elbow method and silhouette analysis are used to determine the optimal number of clusters, resulting in four different clusters. The application of the K-Means algorithm produces the following cluster characteristics: Cluster 1, with good environmental conditions and access to health services, shows low levels of stunting; Cluster 2, with moderate environmental conditions, shows moderate levels of stunting; Cluster 3, which is characterized by poor living conditions and limited access to health services, has levels high stunting; and Cluster 4, with varied environmental conditions but very limited access to health and sanitation services, also shows a high stunting rate. Validation using the Silhouette Coefficient produces an average score of 0.65 which indicates good clustering quality shows that environmental factors, access to health services, and sanitation conditions have a significant impact on the incidence of stunting. Based on these findings, policy and intervention recommendations are focused on Clusters 3 and 4, which have high stunting rates. The interventions carried out include increasing access and quality of nutrition, health services, sanitation conditions, economic empowerment, and health education.

Simple IoT-Based Home Security System Using ESP32 and Blynk

2025-08-27T07:11:42+07:00

Advances in Internet of Things (IoT) technology have brought about various innovations in home security systems. This research aims to design and implement an IoT-based home security system integrated with the ESP32, the KY-037 sound sensor, and the HC-SR501 PIR motion sensor. This system is capable of detecting suspicious movement or sounds around the home door and then sending real-time notifications to the Blynk app on the user's smartphone. The method used was the development of an ESP32-based prototype connected to the internet to transmit sensor data. Test results show that the system can provide notifications with a response time of less than 2 seconds. This system is easy to implement, energy efficient, and suitable for household use. This research demonstrates the great potential of IoT in improving security with affordable costs and high flexibility. The study also highlights the importance of sensor threshold testing to improve detection accuracy. Further developments could include integration with IP cameras and facial recognition systems.

Safety Measures For Special Care Individuals At The Bureau Of Fire Protection In Bolinao, Pangasinan, Philippines:Basis For A Plan Of Action

2025-07-11T15:16:36+07:00

This study assessed the implementation of fire safety measures by the Bureau of Fire Protection (BFP) in Bolinao, Pangasinan, focusing on the needs of special care individuals, including persons with disabilities, the elderly, and those with mobility or cognitive impairments. Using a descriptive-comparative and correlational research design, data were gathered from 150 special care individuals and caregivers, and 20 BFP personnel through validated questionnaires. The results revealed that while special care individuals generally perceived the level of fire safety implementation as high across awareness, preparedness, facilities, training, and policy compliance, BFP personnel rated them more moderately, highlighting gaps in training and infrastructure. Statistical analysis showed significant discrepancies in perception and underscored the need for targeted interventions. The study concludes with a proposed action plan aimed at enhancing inclusive fire safety protocols, training, and equipment for vulnerable populations.

Blockchain Implementation on Subsidised LPG Distribution in Gas Supply Chain (Case Study: Medan)

2024-06-19T14:44:58+07:00

This study examines the potential of private blockchain technology on the Multichain platform for the implementation of a subsidised gas distribution information system in Medan. The objective is to enhance transparency, security, and reliability. The data were collected via a literature review and documentation analysis, and the system was developed using the waterfall methodology. The Multichain-based architecture ensures secure, transparent, and traceable transactions, thereby reducing the incidence of fraud and discrepancies. The results demonstrate that the architecture fulfils the requisite criteria, establishing a robust framework for gas distribution. This validates the effectiveness of Multichain-based private blockchain for improved efficiency and reliability in Medan's subsidised gas distribution system.

Identification Of Malaria Parasites Plasmodium Vivax on Red Blood Cells Using the Probabilistic Neural Network Method

2025-08-21T20:23:17+07:00

Malaria is a disease that is infects human red blood cells transmitted through the bite of a female Anopheles mosquito that contains the parasite genus Plasmodium. Plasmodium vivax is one of the types of parasites that causes malaria, which is known as the type of malaria with the widest distribution area, from tropical, subtropical to cold climates. The diagnosis of malaria, basically depend on microscopic analysis of Giemsa-smeared thin and thick films of blood. However, this diagnostic method is time consuming and prone to human error. To overcome this problem, a method is needed to automatically identify malaria parasites on red blood cells. This study proposes to identifying the malaria parasite Plasmodium vivax using the Probabilistic Neural Network method. The steps taken before identification are preprocessing using Green Channel, Contrast Limited Adaptive Histogram Equalization (CLAHE), Morphological Close and Background Exclusion, then segmentation with Otsu Thresholding, next step is post- processing with Connected Component Analyst (CCA) and feature extraction with Invariant Moment. The results of this research showed that the method used was able to identify the malaria parasite plasmodium vivax on microscopic images of reb blood cells with an accuracy rate of 97.14%, and sensitivity of 95%.