Article Per Year (5 Year)
p-Index From 2020 - 2025
2.217
P-Index
This Author published in this journals
All Journal KOPERTIP: Jurnal Ilmiah Manajemen Informatika dan Komputer Jurnal Ilmiah Betrik : Besemah Teknologi Informasi dan Komputer Jurnal Informatika dan Rekayasa Perangkat Lunak Narada : Jurnal Desain dan Seni Jurnal Informatika Terpadu Jurnal Ilmiah Betrik : Besemah Teknologi Informasi dan Komputer
Umi Hayati
STMIK IKMI Cirebon
Arts Computer Science & IT Decision Sciences, Operations Research & Management Education

Published : 14 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 5 Documents
Search
Journal : Jurnal Informatika dan Rekayasa Perangkat Lunak

 1 
Klasifikasi Tumor Otak menggunakan Convolutional Neural Network dan Transfer Learning Muhammad Hasan Fadlun; Martanto Martanto; Umi Hayati
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 1 (2024): Maret
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i1.10318

Abstract

Brain tumor is an uncontrolled growth of cells in the form of a mass or tissue within the brain, capable of producing both cancerous and non-cancerous symptoms. Brain tumors are part of a group of tumors involving the nervous system, including tumors in the spinal cord and peripheral nerves. It is not a common disease, and prompt intervention is necessary to receive timely medical treatment or appropriate therapy. This research aims to apply Deep Learning techniques in the automatic classification of brain tumors. In this study, a dataset of brain MRI images covering various types of brain tumors was used. The dataset consisted of 3264 MRI images with four classes: glioma, meningioma, pituitary, and no tumor, obtained from Kaggle.com. The system utilized a pre-trained CNN architecture, EfficientNet-B0, trained on the ImageNet dataset. In the Transfer Learning phase, fine-tuning was performed on the last layers of the CNN to adapt it to the brain tumor image dataset. The Convolutional Neural Network model was trained using MRI images to identify important features related to brain tumors. Subsequently, with Transfer Learning, the knowledge acquired by the pre-existing model was adopted and applied to a new dataset to enhance model performance. The application of Deep Learning techniques in the automatic classification of brain tumors provides significant benefits in medical practice. With this system, doctors and radiologists can obtain more effective assistance in diagnosis and treatment planning. The ability to automatically recognize brain tumors with high accuracy also enables the adoption of this technology in various medical facilities, thereby improving the accessibility of testing and treatment needed by patients. The results of this research demonstrate that the CNN and TL methods successfully achieved high performance, including an epoch accuracy of 0.9981 or 99%, a loss of 0.0061, and an evaluation with values generated by the confusion matrix showing high precision of 0.98 or 98%, recall of 0.98 or 98%, and an F1-score of 0.98 or 98%. This study illustrates the significant potential of implementing Deep Learning techniques, particularly CNN and TL, in the automatic classification of brain tumors. Advances in this field can contribute significantly to improving the diagnosis, treatment, and prognosis of brain tumor patients, accelerating efforts to address this complex disease.
Analisis Sentimen Komentar Pengguna Youtube terhadap Kebijakan Baru Badan Penyelenggara Jaminan Kesehatan Sosial Menggunakan Naïve Bayes Muhamad Taufik Sugandi; Martanto Martanto; Umi Hayati
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 1 (2024): Maret
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i1.10301

Abstract

Many social media platforms are used by the public to express opinions and seek information. YouTube is a media sharing site, a kind of virtual entertainment for sharing video and audio media. YouTube has become one of the most popular video viewing platforms today. There are various topics discussed in YouTube videos, one of which is the discussion about the new policy of removing class 1, 2, and 3 systems and replacing them with the Standard Inpatient Class (KRIS) system in the Social Security Administrator (BPJS) for Health. Health is also a very important issue and is still a topic that is frequently discussed everywhere and anytime. BPJS for Health greatly helps the public in overcoming the declining economy, with the existence of BPJS for Health the public does not need to pay for medical expenses. Therefore, sentiment analysis will be conducted on the services provided by BPJS for Health to determine whether public opinion about BPJS is positive, neutral, or negative. The algorithm used is Naïve Bayes. In this sentiment analysis, 2,968 datasets were crawled from YouTube using several keywords related to BPJS for Health. Based on the research results using the Naïve Bayes algorithm, the highest accuracy of the model on the test data reached 96% with a ratio of 80:20. This indicates that the model is capable of classifying sentiment in comments well. This study is dominated by positive sentiment comments at 45.9% or 1,354 data out of a total of 2,948 comment data, indicating strong support for the new policy and many who are very helped by the services of BPJS for Health.
Analisis Sentimen Review Hotel Menggunakan Metode Naïve Bayes pada Hotel di Wilayah Kota Cirebon Muhamad Jihad Andiana; Martanto Martanto; Umi Hayati
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 1 (2024): Maret
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i1.10312

Abstract

Cirebon, a city in West Java, Indonesia, is known for its various tourist attractions, including culinary and historical sites. However, finding the right accommodation can be a challenge. To address this issue, a study has analyzed 875 hotel reviews in Cirebon from Google Maps, using the Naive Bayes method and the TF-IDF algorithm. The aim of this study is to help tourists get a better picture in choosing a hotel. The results show that this algorithm successfully achieved an accuracy of 90.52% in identifying whether the review was positive or negative. Even without the use of the SMOTE operator, the accuracy remains high, at 75.66%. So, this study provides a data-based solution for choosing a hotel in Cirebon.
Clustering Status Gizi Balita menggunakan Metode K-Means pada Posyandu Desa Mekar Wangi Muhamad Djaelani; Martanto Martanto; Umi Hayati
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 1 (2024): Maret
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i1.10321

Abstract

The health of children under five is very important in the development of a country. Toddler nutrition is a key aspect in ensuring the healthy growth and development of children. This study aims to analyze the clustering of nutritional status of toddlers in Mekar Wangi village using the K-Means algorithm. Clustering analysis is a data mining analysis method that is influenced by the clustering algorithm method. The nutritional status of toddlers at the posyandu in Mekar Wangi Village is grouped based on certain metrics, such as body weight and height, using the K-Means Clustering technique. Data contains a lot of attribute information. Once the data is collected and analyzed, pre-processing is performed to remove invalid and empty data. The results of the clustering analysis show that some groups of toddlers have normal nutritional status, while other groups have less or more nutritional problems. The optimal Davies Bouldin Index (DBI) performance evaluation value was found using the RapidMiner tool with K2 and the value of 0.164 which is close to 0 indicates that the evaluated cluster produced a good cluster. With a better understanding of the nutritional patterns of toddlers in Mekar Wangi Village, Posyandu officers can developing a more efficient program to improve the nutritional quality of children in Mekar Wangi Village. Posyandu officers can assist in decision making to develop more targeted recommendations and interventions to improve the nutritional status of toddlers in Mekar Wangi village.
Analisis Sentimen Pengguna Youtube terhadap Polemik Pelarangan Tiktok Shop menggunakan Algoritma Naive Bayes Muhamad farhan Tholhah hidayat; Martanto Martanto; Umi Hayati
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 1 (2024): Maret
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i1.10313

Abstract

Youtube and TikTok are creative platforms for creating videos and interacting with users. In addition to its function as a creative platform, TikTok Shop has recently emerged as a new breakthrough in the world of e-commerce because it can combine social media and e-commerce in one platform. TikTok Shop has become controversial as it disrupts micro, small, and medium-sized enterprises (MSMEs). Due to this controversy, the Indonesian government, through the Ministry of Home Affairs under the instruction of the President of Indonesia, has officially prohibited the use of TikTok as an e-commerce platform and limited it to only being a social media or social commerce application, leading to controversy turning into polemics. This has elicited various reactions from TikTok users, MSMEs, the general public, sellers, and TikTok Shop customers. Therefore, a method is needed to classify reviews automatically by conducting sentiment analysis. In this study, 4403 comment data from one CNN YouTube content titled 'TikTok Shop Banned? Ministry of Cooperatives and SMEs: If Not Regulated, Our MSMEs Could Collapse' were collected. This research applied the naïve Bayes algorithm with a qualitative and quantitative integration method and used the Knowledge Discovery in Databases (KDD) approach and confusion matrix evaluation. The data were divided into training and test sets using four schemes: first scheme 90-10, second scheme 80-20, third scheme 70-30, and fourth scheme 60-40. After evaluating the third scheme with a 70-30% data split, it achieved the best accuracy with a 94% accuracy rate of the test data in the naïve Bayes confusion matrix, which is the percentage of successfully predicted data. Furthermore, the Recall value was 96%, Precision 98%, and F1-Score 96%. This indicates that the model has a high level of accuracy for all training and test data.
Co-Authors Arif Rinaldi Dikananda Cep Lukman Rohmat Dessy Angelina Dodi Solihin Edi Wahyudin Fathurrohman Gifthera Dwilestari Heliyanti Susana Iin Kaslani Martanto . Mohamad Ripai Muhamad Djaelani Muhamad farhan Tholhah hidayat Muhamad Jihad Andiana Muhamad Taufik Sugandi Muhammad Hasan Fadlun Muhammad Saifurridho Mulyawan Nova Nur Sakbani Puspita Maulana Arumsari Rahayu Arumsari Rosdiana Rosdiana Ruli Herdiana Ruli Herdiana Ryan Hmonangan Shalsa Fadillah Rahmadenta siti azhar Tati Suprapti Wita Widyawati