Article Per Year (5 Year)
p-Index From 2021 - 2026
1.037
P-Index
This Author published in this journals
All Journal JURNAL MEDIA INFORMATIKA BUDIDARMA Jurnal Sisfokom (Sistem Informasi dan Komputer) TEKNOLOGIA Building of Informatics, Technology and Science Journal of Data Science and Its Applications Jurnal Sistem Komputer dan Informatika (JSON) eProceedings of Engineering
Dody Qori Utama
Telkom University
Automotive Engineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering Industrial & Manufacturing Engineering Mathematics

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

Found 14 Documents
Search

 1   2 
Identifikasi Citra berdasarkan Gigitan Ular menggunakan Metode Active Contour Model dan Support Vector Machine Dewangga, Dhiya Ulhaq; Adiwijaya, Adiwijaya; Utama, Dody Qori
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 3, No 4 (2019): Oktober 2019
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v3i4.1409

Abstract

Tropical countries have a warm and humid climate are suitable habitat for the lives of reptile animals, especially snakes. Snakes are a type of reptile animal that is widely found in tropical countries, especially in Indonesia. The worst thing that happens when meeting a snake is the bite of snake. If the bite comes from a venomous snake it can cause a more serious problem than the bite from non-venomous snake is, which can cause paralysis, disability, and the worst is death. According to the WHO (World Health Organization) an estimated 5.4 million people are bitten by snakes each year with almost 2.7 million being bitten by venomous snakes and get affected symptoms. Around 81,000 to 138,000 people die every year. This research uses image processing technic to make the identification system of snake bites whether venomous or non-venomous. The method used in this system is Active Contour Model and Support Vector Machine. By using these methods, the highest accuracy is obtained in the best of SVM kernel, on RBF kernel and Polynomial kernel.
Classification of Electrocardiogram Signals using Principal Component Analysis and Levenberg Marquardt Backpropagation for Detection Ventricular Tachyarrhythmia Astrima Manik; Adiwijaya Adiwijaya; Dody Qori Utama
Journal of Data Science and Its Applications Vol 2 No 1 (2019): Journal of Data Science and Its Applications
Publisher : Telkom University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21108/jdsa.2019.2.12

Abstract

Abstract Ventricular Tachyarrhythmia (VT) are the primary arrhythmias which are cause of sudden death. For someone who already has symptoms of VT should immediately perform an examination of one of them by using an electrocardiogram (ECG). An electrocardiogram is a recording of the heart's electrical results in a waveform. However, limited ability in analysis and diagnosis of ECG reading is still difficult to do. Therefore, the classification of ECG signals is needed to detect a person, especially those with VT or not. In this research focuses on the classification of VT heartbeats from ECG signals by using median filter method in preprocessing, Principal Component Analysis (PCA) as feature extraction and modified Backpropagation (MBP) as classification. This research used machine learning method that is a neural network with backpropagation modification that is Levenberg Marquardt to speed up network training process. The best VT detection performance results were based on the average accuracy of the overall scheme of 91.67% with the best parameters that principal component=10 and 20, hidden neuron=4, and µ value=0.001 as well training time 1 seconds with a comparison of train data and test data that is 80:20 percent. Keywords: Electrocardiogram, Levenberg Marquardt Backpropagation, Median filter, Principal Component Analysis, and Ventricular Tachyarrhythmia
Snakebite Classification Using Active Contour Model and K Nearest Neighbor Chiara Janetra Cakravania; Dody Qori Utama
Journal of Data Science and Its Applications Vol 3 No 1 (2020): Journal of Data Science and Its Applications
Publisher : Telkom University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34818/jdsa.2020.3.38

Abstract

Indonesia is categorized as one of tropical countries that have a high risk of snakebites. This surely may endanger rural citizens’ lives for there are still many snakes found in rural areas. The main cause of death from snakebite cases is by reason of the venom squirted from snake’s canine teeth. Others causes are errors in identifying the bite marks visually. There are anatomical differences between puncture wounds from venomous and non-venomous snakes. This study established a snakebite identification system using Active Contour Model and K Nearest Neighbor (KNN) methods. By performing some tests related to the parameters used in the method, the highest accuracy value on K Nearest Neighbor method was obtained by using the correlation distance rule, the K value = 3, without using distance weight in the classification system.
Klasifikasi Gambar Gigitan Ular Menggunakan Regionprops dan Algoritma Decision Tree Yoga Widi Pamungkas; Adiwijaya Adiwijaya; Dody Qori Utama
Jurnal Sistem Komputer dan Informatika (JSON) Vol 1, No 2 (2020): Januari 2020
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (408.081 KB) | DOI: 10.30865/json.v1i2.1789

Abstract

Indonesia has a high biodiversity of snakes. Snake species that exist throughout Indonesia, consisting of venomous and non-venomous snakes. One of the dangers that can be posed by snakes is the bite of several types of deadly snakes. Snake bite cases recorded in Indonesia are quite high with not a few fatalities. Most of the deaths caused by snakebite occur due to errors in the handling procedure for the bite wound. This problem can be overcome one of them if we know how to classify snake bite wounds, whether venomous or non-venomous. In this study, a classification system for snake bite wound image was built using Regionprops feature extraction and Decision Tree algorithm. Snake bite images are classified as either venomous or non-venomous without knowing the kind of the snake. In Regionprops several features are used to help the process of feature extraction, including the number of centroids, area, distance, and eccentricity. Evaluation of the model that was built was found that the parameters of the number of centroids and the distance between centroids had the most significant influence in helping the classification of images of snakebite wounds with an accuracy of 97.14%, precision 92.85%, recall 91.42%, and F1 score 92.06%.
KLASIFIKASI GIGITAN ULAR MENGGUNAKAN LOCAL BINARY PATTERN DAN NAÏVE BAYES Fathur Rohman; Adiwijaya Adiwijaya; Dody Qori Utama
JURNAL TEKNOLOGIA Vol 2 No 1 (2019): Jurnal Teknologia
Publisher : Aliansi Perguruan Tinggi Badan Usaha Milik Negara (APERTI BUMN)

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (5494.817 KB)

Abstract

Cases of poisonous snake bites around the world are estimated to occur around 421,000 cases and 20,000 of them die every year. Identifying snake bite marks on victims will greatly help the medical team in handling victims of snake bites and will avoid fatal errors such as the death of the victim. This research will try to create a system that can classify snake bites images. The system has been built using the extraction method Local Binary Pattern (LBP) and Naive Bayes. Parameter r is a radius, while paramter P is the number of neighbor . The best result of this system has accuracy 83.33%, precision 1.00, recall 0.75, and F1 Score 0.86,parameter that used are r = 1 with P = 8 and r = 3 with P = 16. The dataset used has 20 data, the data divided into 14 training data and 6 testing data.
Smartphone Purchase Recommendation System Using the K-Nearest Neighbor (KNN) Algorithm Bayu Rahmat Setiaji; Dody Qori Utama; Adiwijaya Adiwijaya
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 6, No 4 (2022): Oktober 2022
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v6i4.4753

Abstract

Indonesia is in the fourth position of the countries with the most smartphone users worldwide. Smartphones are needed in today's modern times. Smartphones are also used not only for long-distance communication but also for carrying out daily work. Smartphones are currently used for study and work and also become entertainment to play. Therefore, smartphones are very much sought after for the suitability of users who carry out their daily activities. So this research is very helpful for users to find smartphones that support their daily activities such as studying, working, and playing. This research is based on a website that can make it easier for users to see their smartphone recommendations directly. The analysis uses the K-Nearest Neighbor (KKN) method to see the ratings reviewed by other users who have tried using their smartphones with different phone brands. The calculation method in the current study uses 3 KNN calculations and uses the concept of combining calculations to find the maximum recommendation results. The result of the recommendation system using the K-Nearest Neighbor method is in the form of a review stating whether the user agrees or disagrees. In the current study, there have been 100 reviews from users, and it has a percentage of 78% for success and 22% for failure.
Klasifikasi Penyakit Aritmia Melalui Sinyal Elektrokardiogram (ekg) Menggunakan Metode Local Features Dan Support Vector Machine Gilang Titah Ramadhani; Adiwijaya Adiwijaya; Dody Qori Utama
eProceedings of Engineering Vol 5, No 1 (2018): April 2018
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Jantung merupakan organ terpenting dalam tubuh manusia dan selalu dituntut dalam keadaan baik, tidak dapat dipungkiri bahwa seseorang memiliki kemungkinan menderita penyakit jantung aritmia. EKG merupakan salah satu cara untuk mendeteksi penyakit jantung. Dengan menggunakan metode Local Features yang merupakan metode ekstraksi ciri dengan menghitung jumlah detak jantung sehingga dapat membantu dalam proses klasifikasi yang dilakukan oleh Support Vector Machine (SVM). Dalam fitur pengklasifikasian yang dilakukan oleh SVM, mendapatkan hasil akurasi dari dua dataset yang digunakan. Untuk data EKG normal akurasi terbesar bernilai 67% yang dihasilkan dari SVM kernel linear dan RBF, untuk data EKG aritmia akurasi terbesar bernilai 83% yang dihasilkan oleh kernel linear dan 16% menggunakan kernel RBF. Perbedaan kernel mempengaruhi akurasi pada setiap data bergantung kepada karakteristik setiap data EKG yang digunakan.
Klasifikasi Sinyal Ecg Gagal Jantung Menggunakan Wavelet Dan Jst Propagasi Balik Dengan Modifikasi Gradien Konjugat Dinda Karlia Destiani; Adiwijaya Adiwijaya; Dody Qori Utama
eProceedings of Engineering Vol 5, No 1 (2018): April 2018
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Gagal jantung kongestif (CHF) merupakan salah satu penyakit mematikan di dunia yang terjadi karena adanya kelainan pada otot-otot jantung sehingga jantung tidak dapat memompa darah sesuai kebutuhan tubuh. Sinyal jantung dapat dideteksi dengan menggunakan alat Electrocardiography (ECG). Pada dasarnya, sinyal jantung normal memiliki bentuk yang serupa. Namun, sinyal jantung penderita CHF, memiliki bentuk yang cukup bervariasi pada setiap individu. Hal tersebut dapat menimbulkan suatu permasalahan jika proses ekstraksi dilakukan dengan pencarian local features secara manual. Oleh karena itu, ekstraksi ciri wavelet digunakan pada penelitian ini untuk memetakan frekuensi terhadap waktu. Di samping itu, proses klasifikasi dengan metode JST Propagasi Balik Standar memerlukan waktu yang cukup lama untuk melakukan proses pelatihan. Sehingga, metode JST Modifikasi Propagasi Balik Gradien Konjugat Polak-Ribiere dengan teknik line search diusulkan untuk mempercepat proses pencarian. Pada akhir penelitian, diperoleh ekstraksi Dekomposisi Paket Wavelet pada level 5 dengan data pelatihan yang digunakan sebanyak 22 menghasilkan nilai rata-rata yang lebih tinggi dari hasil pengujian lainnya, yaitu sebesar 72.5%. Adapun jumlah neuron yang paling optimal untuk digunakan pada lapisan tersembunyi yaitu sebanyak 30 neuron. Sementara itu, Charalambous’ Search merupakan teknik pencarian yang tercepat dan terakurat untuk diterapkan pada kasus ini dengan waktu pencarian 2.65 detik, 14 epoch, serta akurasi 87.5%.
Pelabelan Sinyal Elektrokardiogram (ekg) Pada Klasifikasi Fibrilasi Atrium Menggunakan Hidden Markov Model Zakia Firdha Razak; Adiwijaya Adiwijaya; Dody Qori Utama
eProceedings of Engineering Vol 5, No 3 (2018): Desember 2018
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Penyakit jantung telah menjadi salah satu penyebab kematian terbesar di dunia. Di antaranya yang paling sering terjadi adalah Fibrilasi Atrium, yaitu kondisi yang meliputi aktivitas sangat cepat dan tidak beraturan dalam atria serta menimbulkan gejala-gejala seperti jantung berdebar, sakit kepala, kehilangan kesadaran, sesak nafas dan rasa letih. Untuk mengenali penyakit ini, akan dikembangkan sebuah sistem pengenalan gelombang EKG. Hidden Markov Model yang menggunakan pendekatan probabilitas telah terbukti menghasilkan performansi yang bagus dalam pengenalan sinyal suara. Oleh karena itu, dalam tugas akhir ini akan diujikan metode HMM dalam kasus elektrokardiogram (EKG). Dalam sistem pendeteksian penyakit jantung terdapat 2 proses yaitu pemodelan dan pengenalan. Pada proses pemodelan akan dibuat suatu model pelabelan sinyal EKG dengan menghitung parameterparameter HMM, yaitu distribusi inisial state, distribusi probabilitas transisi antar state, dan distribusi probabilitas symbol observasi pada suatu state. Setelah diperoleh ketiga nilai tersebut yang optimal, maka terbentuklah suatu model HMM untuk jenis penyakit jantung. Sedangkan proses pengenalan penyakit jantung dilakukan pada tiap data EKG dengan menghitung likelihood dari data testing yang akan dikenali terhadap semua model data EKG yang telah dilatih sebelumnya. Dengan pelabelan yang teliti dan penentuan nilai probabilitas observasi yang optimal, HMM dapat digunakan untuk mengenali penyakit jantung. Hasil dari pengujian menunjukkan, nilai probabilitas observasi yang didapat dari nilai random selalu menghasilkan akurasi yang berubah-ubah, berbeda jika disama ratakan dengan jumlah state dan jumlah pengamatan. Selain itu, perbandingan jumlah state serta proporsi data juga berpengaruh terhadap akurasi. Akurasi terbesar adalah 100% dengan 2 dan 4 state ketika proporsi data training sebanyak 75% dan nilai probabilitas observasi diatur random. Kata kunci : hidden markov model, elektrokardiogram, distribusi inisial state, distribusi probabilitas transisi antar state, distribusi probabilitas simbol observasi Abstract Heart disease has become one of the biggest causes of death in the world. Among the most common is Atrial Fibrillation, a condition that includes very fast and irregular activities in atria and causes many symptoms such as palpitations, headaches, loss of consciousness, shortness of breath and tiredness. To recognize this disease, an ECG wave recognition system will be developed. Hidden Markov Model that use a probability approach has been proven to make good performance in speech signal recognition. Therefore, in this final project the HMM method will be tested on electrocardiogram (ECG) signal. In the detection system for heart disease there are 2 processes; modeling and recognition. In the modeling process, ECG signal labeling model will be made by calculating HMM parameters, namely the initial state distribution, the state transition probability distribution, and the observation symbol probability distribution. After obtaining these three optimal values, an HMM then formed for each type of heart disease. The recognition process of heart disease is carried out on each ECG data by calculating likelihood from testing data that will be identified by all models of ECG record after they have been previously trained. With appropiate labeling and determination of optimal observation probability values, HMM can be used to identify heart disease. The results showed that the probability value of the observations obtained from random values always results in varying accuracy. The comparison of the state number and the data proportion also affects accuracy. The highest accuracy is 100% with 2 states when the proportion of training data is 75% and the probability value is arranged randomly. Keywords: hidden markov model, elektrocardiogram, initial state distribution, state transition probability distribution, observation symbol probability distribution
Klasifikasi Bekas Gigitan Ular Menggunakan Active Contour Model Dan K Nearest Neighbor Chiara Janetra Cakravania; Adiwijaya Adiwijaya; Dody Qori Utama
eProceedings of Engineering Vol 6, No 2 (2019): Agustus 2019
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

AbstrakIndonesia termasuk salah satu negara tropis yang memiliki tingkat risiko terkena gigitan ular yang tinggi.Penyebab utama kematian dari kasus gigitan ular tentunya adalah karena racun yang disemprotkan olehular berbisa tersebut melalui gigi taringnya. Penyebab lainnya adalah kesalahan dalam mengidentifikasibekas gigitan tersebut secara visual. Terdapat perbedaan anatomi gigitan pada ular berbisa dan tidaksehingga mengakibatkan perbedaan bekas gigitan pada korban. Pada penelitian ini dibangun sistemidentifikasi bekas gigitan ular yang dapat mengidentifikasi bekas gigitan ular tersebut oleh ular berbisaatau tidak dengan metode Active Contour Model dan K Nearest Neighbor. Dengan melakukan beberapapengujian terkait parameter yang digunakan pada metode tersebut, didapat nilai akurasi tertinggi padametode K Nearest Neighbor adalah dengan menggunakan aturan jarak correlation, nilai K = 3, dan sistemklasifikasi tidak menggunakan distance weight. Kata kunci : bite mark, active contour, k nearest neighborAbstractIndonesia is categorized as one of tropical countries that have a high risk of snakebites. This surely mayendanger rural citizens’ lives for there are still many snakes found in rural areas. The main cause of deathfrom snakebite cases is by reason of the venom squirted from snake’s canine teeth. Others causes areerrors in identifying the bite marks visually. There are anatomical differences between puncture woundsfrom venomous and non-venomous snakes. This study established a snakebite identification system usingActive Contour Model and K Nearest Neighbor (KNN) methods. By performing some tests related to theparameters used in the method, the highest accuracy value on K Nearest Neighbor method was obtainedby using the correlation distance rule, the K value = 3, without using distance weight in the classificationsystem. Keywords: bite mark, active contour, k nearest neighbor
Co-Authors Adiwijaya Aji Gautama Putrada Astrima Manik Bayu Rahmat Setiaji Chiara Janetra Cakravania Chiara Janetra Cakravania Dewangga, Dhiya Ulhaq Dinda Karlia Destiani Fathur Rohman Gilang Titah Ramadhani Gunarto, Dzaki Mahadika Jendral Muhamad Yusuf Zia Ul Haq Mahendra Dwifebri Purbolaksono Mbake Woka, Adrian Yoris Muhammad Dafa Prima Aji sa'adah, siti Yoga Widi Pamungkas Zakia Firdha Razak