Garuda - Garba Rujukan Digital

Klasifikasi Suara Anjing Menggunakan Pretrained Model Yet Another Mobile Network Berbasis Convolutional Neural Network Djuardi, Rich Deshan; Rochadiani, Theresia Herlina
Building of Informatics, Technology and Science (BITS) Vol 6 No 1 (2024): June 2024
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i1.5165

In everyday life, pets such as dogs often become an inseparable part of human life. Motivations for keeping a pet can vary from individual to individual, ranging from the need for a loyal companion to the responsibility of caring for another living creature. Among the various choices of pets, dogs are often considered the most loyal and loyal friends towards humans. This uniqueness makes many people choose to keep dogs as part of their family. Often, dog owners may not understand the message that the sounds produced by their beloved pets are trying to convey. These dog sounds have a special purpose that can reflect various emotions, such as joy, sadness, or anger. A dog's voice can also be an indicator of their health that owners need to pay attention to. The main focus of this research is to develop dog voice classification technology to help owners understand and communicate with their pet dogs. In this research, a pre-trained YAMNet model is used as a basis for classifying various audio events. The model training process uses the CNN algorithm contained in the YAMNet architecture. The total data used was 373 data which were classified into 4 classes, namely, bark, howling, growling, whimper. The results of this research model achieved 97.8% accuracy with precision, recall and f1-scores for each class >= 95%.

Design of Batak Toba Script Recognition System Using Convolutional Neural Network Algorithm Steven Willian; Rochadiani, Theresia Herlina; Thamrin Sofian
Sinkron : jurnal dan penelitian teknik informatika Vol. 7 No. 3 (2023): Article Research Volume 7 Issue 3, July 2023
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/sinkron.v8i3.12617

Indonesia is one of the countries with diversity and abundant cultural wealth, one of which is the Batak Toba script as one of the wealth originating from the Batak tribe. However, the existence of the Batak Toba script is decreasing along with the rapid development of the times, due to the lack of interest of the younger generation and public awareness in preserving the Batak Toba script. From these problems, the author conducted research to create a model of introducing the Batak Toba script, as an effort to preserve the Batak Toba script which is one of Indonesia's cultural wealth. The purpose of this research is to create a Batak Toba script recognition model using a digital handwriting dataset, and has an output in the form of visual text and with audio pronunciation of each script. The method used in this research is the Convolutional Neural Network algorithm combined with RMSprop optimizer. Convolutional Neural Network is an algorithm that is one of the deep learning methods that has good performance on image data. The results of this study incised a recognition model with a relatively high level of accuracy, which is equal to 99,54% which was tested on the Batak Toba script dataset in the form of digital handwriting. Through this research, the model using the Convolutional Neural Network algorithm used in this research is able to produce good results for recognizing the Batak Toba script in the form of handwriting.

Deteksi Potensi Menyontek Menggunakan Feedforward Neural Network Pada Ujian Daring Hadibrata, Lymanto; Rochadiani, Theresia Herlina
SINTECH (Science and Information Technology) Journal Vol. 7 No. 2 (2024): SINTECH Journal Edition Agustus 2024
Publisher : Prahasta Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31598/sintechjournal.v7i2.1585

Pendidikan di Indonesia merupakan salah satu faktor pendukung yang dapat menjadikan Indonesia menjadi negara maju. Akan tetapi masih banyak pelajar yang melakukan praktik menyontek sehingga menurunkan kualitas pendidikan di Indonesia. Untuk mengurangi praktik menyontek di Indonesia, penelitian ini bertujuan untuk membuat sebuah model deep learning dengan menggunakan metode feedforward neural network untuk mendeteksi potensi menyontek. Penelitian ini menggunakan 51 video dataset diperoleh dari orang-orang yang pada akhirnya diubah menjadi titik-titik koordinat menggunakan Mediapipe Face Landmark yang disimpan pada file CSV. Pada penelitian ini terdapat 7 class pada dataset yang sudah dibuat yaitu netral, hadap_atas, hadap_bawah, hadap_kiri, hadap_kanan, retina_kiri dan retina_kanan. Indikator utama yang paling menentukan untuk mendeteksi potensi tidak menyontek adalah class netral. Akan tetapi, class retina_kiri dan retina_kanan juga ikut berpartisipasi karena ada pertimbangan dari segi pembacaan soal. Indikator yang menentukan untuk mendeteksi potensi menyontek adalah class hadap_kiri, hadap_kanan, hadap_atas, dan hadap_bawah. Penelitian ini menghasilkan model yang dapat memprediksi potensi menyontek dengan akurasi sebesar 91.6% dengan menggunakan metode feedforward neural network. Dari model yang dihasilkan, dapat diimplementasikan kedalam sebuah sistem ujian daring.

Implementasi Ensemble Deep Learning Untuk Analisis Sentimen Terhadap Genre Game Mobile Cahyadi, Marcelinus Fajar; Rochadiani, Theresia Herlina
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 8, No 3 (2024): Juli 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v8i3.7832

The rapid growth of the online gaming industry in Indonesia has prompted developers to address various challenges in creating successful mobile games. This study aims to evaluate the effectiveness of ensemble learning techniques, particularly soft voting, in enhancing sentiment analysis accuracy across 17 genres of mobile games. Additionally, it identifies the most effective deep learning model for sentiment classification. The research compares the performance of CNN-LSTM, BERT, and CNN-GRU models, as well as an ensemble of these models. Review data was collected from the Google Play Store, then labeled and cleaned to improve data quality, categorized into positive, neutral, and negative sentiments. Data preprocessing techniques included cleaning, case folding, tokenization, normalization, stopword removal, and stemming. Word embedding techniques used were Word2vec for CNN-LSTM and CNN-GRU models, and IndoBERT for BERT model. Ensemble learning combined predictions from these models, significantly improving classification accuracy. Results indicate IndoBERT achieved an accuracy of 89%, while CNN-GRU and CNN-LSTM showed accuracies around 84-85%. The ensemble approach using soft voting successfully increased overall accuracy to 90% by combining predictions from all three models. The study concludes that ensemble learning effectively combines individual model strengths to enhance sentiment classification accuracy. Furthermore, user preference visualization for game genres revealed high popularity for "Strategy", "Word", and "Trivia" genres, while "Sports" genres were less favored. This research is expected to contribute to game developers in determining which genres to develop to enhance success chances and user satisfaction.

Rancang Bangun Alat Pemberi Makan Hewan Peliharaan Pintar Menggunakan Mikrokontroler ESP32 Berbasis Internet Of Things (IoT) dengan Platform Blynk Canady, Richwen; Danendra, Danica Recca; Indrawan, Vincenzo Matalino; Rochadiani, Theresia Herlina
Jutis (Jurnal Teknik Informatika) Vol. 11 No. 2 (2023): Jutis (Jurnal Teknik Informatika)
Publisher : Universitas Islam Syekh Yusuf

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33592/jutis.v11i2.3460

Internet of Things (IoT) dapat digunakan untuk menangani berbagai masalah dalam kehidupan sehari-hari manusia, terutama dalam hal-hal yang memerlukan pengawasan konstan. Salah satu contohnya adalah pemberian makan hewan peliharaan. Penelitian ini bertujuan untuk merancang alat pemberi makan hewan peliharaan pintar berbasis Internet of Things menggunakan mikrokontroler ESP32 dan platform Blynk. Perangkat ini menawarkan fitur pemberian makan otomatis dan manual jarak jauh, serta pemantauan real-time makanan yang tersedia melalui platform dan aplikasi mobile Blynk. Studi sebelumnya tentang pemberi makan hewan peliharaan otomatis telah ditinjau dan menjadi dasar untuk penelitian ini. Perangkat yang diusulkan mencakup komponen seperti sensor ultrasonik, load cell, modul Real Time Clock, dan mikrokontroler ESP32. Arsitektur tiga lapisan, termasuk lapisan persepsi, lapisan jaringan dan gateway, dan lapisan aplikasi, diimplementasikan dalam desain sistem. Diagram Fritzing memberikan representasi visual dari hubungan komponen. Metodologi penelitian melibatkan identifikasi masalah, tinjauan literatur, desain perangkat, implementasi, dan pengujian sistem IoT. Pengujian sensor dan komponen dilakukan untuk memastikan akurasi dan efisiensi. Antarmuka aplikasi mobile dikembangkan menggunakan platform Blynk, memungkinkan pengguna memantau dan mengontrol pemberi makan pintar secara jarak jauh. Secara keseluruhan, perangkat pemberi makan hewan peliharaan pintar menunjukkan kinerja yang dapat diandalkan dan berhasil mengatasi tantangan yang terkait dengan pemberian makan hewan peliharaan.

Sentiment Analysis of YouTube Comments Toward Chat GPT Rochadiani, Theresia Herlina
Jurnal Transformatika Vol 21, No 1 (2023): July 2023
Publisher : Jurusan Teknologi Informasi Universitas Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26623/transformatika.v21i2.7033

Sentiment analysis is used for analyzing the emotions and attitudes expressed in text data. In this study, sentiment analysis is used to understand people’s enthusiasm toward Chat GPT. The primary objective of this study is to investigate the acceptance of people of new artificial intelligence technology, Chat GPT, that may change the future. To get a deep understanding of it, a large dataset of user comments from YouTube is collected and then data pre-processing is done by removing stop words, punctuations, and irrelevant information. Using Text Blob and VADER approaches, comments are classified into positive, neutral, and negative categories. The result shows that most users have a positive sentiment to receive and use Chat GPT. The contribution of this study is to provide insights into the sentiment of people’s response to Chat GPT, which can inform user acceptance of the language model development and give guide its future applications.

Implementation of K-Means, Hierarchical, and BIRCH Clustering Algorithms to Determine Marketing Targets for Vape Sales in Indonesia Laurenso, Justin; Jiustian, Danny; Fernando, Felix; Suhandi, Vartin; Rochadiani, Theresia Herlina
Journal of Applied Informatics and Computing Vol. 8 No. 1 (2024): July 2024
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v8i1.4871

In today's era, smoking is a common thing in everyday life. Along with the development of the times, an innovation emerged, namely the electric cigarette or vape. Electric cigarettes or vapes use electricity to produce vapor. The e-cigarette business is very promising in today's business world due to the consistent increase in market demand. However, determining the target buyer is one of the things that is quite important in determining the success of a business. In this analysis, the background of each region in Indonesia has different diversity; therefore, observation of data is needed to find out which regions in Indonesia have the potential to increase marketing based on profits (margins) to support the target market analysis process so that companies do not suffer losses and increase business success. In this study, the analysis will be carried out using vape quantity, margin, and purchasing power data in each region, which is processed using 3 algorithms: K-Means, Hierarchical, and BIRCH. The results of the clustering of the three algorithms produce two clusters. The K-means, Hierarchical, and BIRCH algorithms produce the same clusters: a potential cluster consisting of 18 cities and a non-potential cluster consisting of 45 cities. To see the performance of the model results, an evaluation was carried out using the Silhouette score, Davies Bouldin, Calinski Harabasz, and Dunn index, which obtained results of 0.765201, 0.376322, 315.949434, and 0.013554. From these results, it can be concluded that the clustering results are not too good and not too bad because the greater the Silhouette Score, Calinski Harabasz, and Dunn Index value, the better the clustering results while for Davies Bouldin the smaller the value means the better the clustering results.

Sistem Navigasi dan Rekomendasi Buku Perpustakaan Berbasis Augmented Reality Karyadi, Phance; Rochadiani, Theresia Herlina; Sofian, Thamrin
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 4 No. 1: MARET 2024
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v4i1.229

Perpustakaan sering menjadi tujuan kunjungan di kampus bagi pengunjung yang ingin mencari dan membaca buku. Waktu yang dibutuhkan untuk melakukan pencarian buku di perpustakaan cenderung lama. Selain itu, untuk menemukan informasi yang serupa dengan kebutuhan pengunjung perlu membaca seluruhnya untuk mengetahui buku yang paling tepat. Studi ini bertujuan untuk mengembangkan sistem navigasi dan rekomendasi buku berbasis augmented reality dengan metode markerless. Penelitian menggunakan pendekatan metode kuantitatif dengan metode pengembangan perangkat lunak prototype. Aplikasi dibangun menggunakan Unity dan ARCore untuk dukungan augmented reality, Immersal SDK untuk lokalisasi dan pemetaan, sistem navigasi menggunakan algoritma A* dan sistem rekomendasi menggunakan TF-IDF dan Cosine Similarity. Berdasarkan hasil pengujian 30 responden diperoleh tujuan ke perpustakaan 3 tertinggi, yaitu mencari buku referensi, membaca buku, dan tempat untuk mengerjakan tugas. Pengalaman mencari judul dan buku sejenis dengan cara konvensional kesulitan dan membutuhkan waktu yang lama. Sebanyak 100% responden setuju aplikasi PraditaLibNav dapat memudahkan untuk mencari judul buku dan mencari rekomendasi buku sejenis. Hasil analisis kuantitatif menggunakan System Usability Scale diperoleh rata-rata skor 78,42 sehingga sistem yang dikembangkan dapat diterima.

Prediction of Air Quality Index Using Ensemble Models Rochadiani, Theresia Herlina
Journal of Applied Informatics and Computing Vol. 8 No. 2 (2024): December 2024
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v8i2.8532

The impact of air pollution on health is measured by the Air Quality Index (AQI). Accurate AQI prediction is essential for pollution reduction and public health recommendations. Traditional methods of monitoring air quality are inaccurate and time-consuming. This study uses IoT-based air quality data from Kampung Kalipaten, Tangerang to build an AQI prediction model with machine learning, specifically an ensemble model. Ensemble techniques such as bagging and boosting, which increase the reliability of predictions by reducing model bias and inconsistency, improve AQI prediction. Four ensemble models used in this study, they are Random Forest Regressor, Gradient Boosting Regressor, Adaboosting Regressor, and Bagging Regressor. As the evaluation, RMSE and R2 metrics used. Random Forest Regressor perform the best with RMSE value of 0.6054 and R2 value of 0.6271, although no significant differences of RMSE and R2 value of the rest models.

Implementasi Ensemble Deep Learning Untuk Analisis Sentimen Terhadap Genre Game Mobile Cahyadi, Marcelinus Fajar; Rochadiani, Theresia Herlina
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 8, No 3 (2024): Juli 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v8i3.7832

The rapid growth of the online gaming industry in Indonesia has prompted developers to address various challenges in creating successful mobile games. This study aims to evaluate the effectiveness of ensemble learning techniques, particularly soft voting, in enhancing sentiment analysis accuracy across 17 genres of mobile games. Additionally, it identifies the most effective deep learning model for sentiment classification. The research compares the performance of CNN-LSTM, BERT, and CNN-GRU models, as well as an ensemble of these models. Review data was collected from the Google Play Store, then labeled and cleaned to improve data quality, categorized into positive, neutral, and negative sentiments. Data preprocessing techniques included cleaning, case folding, tokenization, normalization, stopword removal, and stemming. Word embedding techniques used were Word2vec for CNN-LSTM and CNN-GRU models, and IndoBERT for BERT model. Ensemble learning combined predictions from these models, significantly improving classification accuracy. Results indicate IndoBERT achieved an accuracy of 89%, while CNN-GRU and CNN-LSTM showed accuracies around 84-85%. The ensemble approach using soft voting successfully increased overall accuracy to 90% by combining predictions from all three models. The study concludes that ensemble learning effectively combines individual model strengths to enhance sentiment classification accuracy. Furthermore, user preference visualization for game genres revealed high popularity for "Strategy", "Word", and "Trivia" genres, while "Sports" genres were less favored. This research is expected to contribute to game developers in determining which genres to develop to enhance success chances and user satisfaction.

Title

Found 20 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 20 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 20 Documents
Search