<span lang="EN-US">The problem of poverty is a scourge for every developing country coupled with the economic crisis that occurred during the COVID-19 pandemic. The impact of these problems is felt directly by the people in Indonesia, especially in the Province of West Sumatra. This study aims to predict and classify the level of poverty status by developing an analytical model based on the deep learning (DL) approach. The methods used in this study include the K-Means method, artificial neural network (ANN), and support vector Machine (SVM). The analytical model will be optimized using the Pearson Correlation (PC) method to measure the accuracy of the analysis. The variable indicator uses the parameters of population (</span><em><span lang="EN-US">X<sub>1</sub></span></em><span lang="EN-US">), poverty rate (</span><em><span lang="EN-US">X<sub>2</sub></span></em><span lang="EN-US">), income (</span><em><span lang="EN-US">X<sub>3</sub></span></em><span lang="EN-US">), and poverty percentage (</span><em><span lang="EN-US">X<sub>4</sub></span></em><span lang="EN-US">). The results of the study present prediction and classification output with a validity level of accuracy of 99.8%. Based on these results, it can be concluded that the proposed DL analysis model can present an updated analytical model that is quite effective in carrying out the prediction and classification process. The research findings also contribute to the initial handling of the problem of poverty.</span>
Teknologi kecerdasan buatan yang berfokus penggunaan data dan algoritma untuk meniru cara manusia belajar, salah satu model pembelajaran tersebut adalah adanya performa komputasi dengan menggunakan teknik deep learning. Convolutional neural network (CNN) adalah arsitektur deep learning yang sering digunakan untuk mengatasi masalah klasifikasi gambar. Selain itu, penggunaan teknik pre-processing pada data dapat membantu meningkatkan performa model dengan memperkaya variasi data. Batik telah menjadi warisan budaya turun temurun di seluruh Indonesia khususnya di daerah Sumatera Barat. Banyaknya pola dan motif batik mengakibatkan sulitnya masyarakat mengidentifikasi motif pada batik. Tujuan penelitian ini adalah mengetahui apakah CNN dapat digunakan untuk klasifikasi batik tanah liat Sumatera Barat. Data yang digunakan Penelitian ini adalah 400 citra batik dan dibagi menjadi 4 kelas, ditentukan 320 citra sebagai data latih dan 80 citra sebagai data uji. Hasil pengujian didapat nilai akurasi sebesar 50%. Tingkat akurasi ini cukup baik sebagai rujukan dalam membangun real application pengenalan motif batik secara umum. Hasil ini menunjukkan metode CNN dapat diterapkan untuk mengklasifikasi batik tanah liat Sumatera Barat.
Mastoiditis occurs due to inflammation that can affect the structure of the mastoid bone. The mastoid bone consists of the mastoid air cell system (MACS) which protects the ear structures and regulates air pressure in the ear and has different sizes and characteristics, making it very difficult to identify precisely. This study aims to identify and find the right MACS size by developing an automatic identification model and obtaining the optimal threshold value in the segmentation process using the extended adaptive threshold (eAT) method. The research dataset uses computed tomography (CT)-scan images of 308 slices of 12 patients indicated for mastoiditis. The results of this study provide identification that has the right MACS accuracy and size. Overall, the optimal segmentation process obtained the smallest threshold value of 57 and the largest threshold value of 63, the smallest MACS size is 4.025 cm2 and the largest is 8.816 cm2 with an accuracy rate of 93.4%. The smaller MACS size indicates inflammation in the mastoid area and these patients require more intensive treatment.
The mastoid air cell system (MACS) protects the structures in the ear and regulates air pressure in the ear cavity. MACS segmentation is very difficult because of the many overlapping object characteristics in the temporal bone. This study aims to accurately identify and measure MACS areas from CT-scan images of mastoiditis patients. The data tested consisted of 128 CT images from 13 different patients. Images were taken using the Siemens SOMATOM Perspective CT Scanner model 10662260 axially. The extended Adaptive Threshold (eAT) method was developed to produce optimal threshold values for each test image. Furthermore, the eAT results are used to convert the test image into a binary image and then applied to the identification and extraction model automatically for reconstruction from 2D to 3D images. Smaller MACS sizes indicate inflammation of the mastoid bone and require intensive care. Thus, this research can be used to help doctors make the right decisions in carrying out further medical actions.
Minangkabau language (ML) is one of the daily communication tools used by the people of West Sumatra, Indonesia. ML is a challenge in communicating. The ML language translation process is necessary to facilitate communication. This study aims to build a translation system for ML into Indonesian by developing the concept of natural language processing (NLP). NLP development adopts the performance of morphology-based Minangkabau language stemming algorithm (MLSA) which can separate basic words with affixes and endings. The research dataset adopts 600 basic ML words sourced from the big Minangkabau dictionary. The results of this study provide analytic output that can translate ML into Indonesian well. These results are presented based on the testing process on basic word input with an accuracy rate of 97.16% and based on text documents of 91.65%. Thus, the MLSA performance process presents the accuracy of the translation process. Based on these results, this research contributes to developing a stemming algorithm model in carrying out the process of removing prefixes, inserts, and suffixes in the Minangkabau language. Overall, this research can be useful as a tool for translating the ML into Indonesian.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.