Articles
ANALISIS METODE RBF-NN DENGAN OPTIMASI ALGORITMA GENETIKA PADA PERAMALAN MATA UANG EUR/USD
Nengah Widiangga Gautama;
Agus Dharma;
Made Sudarma
Jurnal Teknologi Elektro Vol 15 No 2 (2016): (July - December) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
Penelitian ini membahas tentang peramalan EUR/USD menggunakan metode RBF-NN (Radial Basis Function – Neural Network) tanpa optimasi dan RBF-NN yang dioptimasi dengan 3 model AG/AGA (Algoritma Genetika dan Algoritma Genetika Adaptif). Sistem RBF-NN dapat diterapkan pada data dengan karakteristik nonlinear dan fluktuatif seperti data EUR/USD. Permasalahan akurasi muncul jika terjadi solusi lokal dalam sistem RBF-NN dan metode AG/AGA dapat digunakan untuk mengatasi solusi lokal tersebut. Keakuratan dari peramalan ditunjukkan lewat nilai MAPE (Mean Absolut Percentage Error). Pada data daily low, metode terbaik adalah Algoritma Genetika II dengan MAPE sebesar 0,2286, sementara pada data daily high metode terbaik adalah Algoritma Genetika Adaptif II dengan MAPE sebesar 0,2190. Metode AG II dan AGA II didukung teknik pencarian di dekat bobot RBF-NN yang terbukti efektif pada kasus mata uang EUR/USD. Perbaikan akurasi yang diberikan AG II dan AGA II terhadap metode RBF-NN dapat diterapkan pada peramalan mata uang lainnya.
Implementasi Algoritma K-Nearest Neighbor pada Perangkat Lunak Pengelompokan Musik untuk Menentukan Suasana Hati
Gede Harsemadi;
Made Sudarma;
Nyoman Pramaita
Jurnal Teknologi Elektro Vol 16 No 1 (2017): (January - April) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (845.53 KB)
Musik erat kaitannya dengan psikologi manusia, kenyataan ini mengindikasikan bahwa musik dapat terkait dengan emosi dan mood/ suasana hati tertentu pada manusia. Setiap musik yang telah tercipta memiliki mood tersendiri yang terpancar, maka dari itu telah banyak penelitian dalam bidang Music Information Retrieval (MIR) yang telah dilakukan untuk pengenalan mood terhadap musik. Penelitian ini menghasilkan sebuah perangkat lunak untuk mengelompokan musik terhadap suasana hati dengan menggunakan algoritma K-Nearest Neighbor. Sistem menerima masukan data berupa file musik format mono *.wav, yang selanjutnya melakukan proses pengelompokan terhadap musik dengan mengggunakan klasifikasi K-NN. Sistem menghasilkan keluaran berupa label jenis mood yaitu, contentment/ kepuasan, exuberance/ gembira, depression/ depresi dan anxious/ cemas; kalut. Secara umum hasil akurasi sistem dengan menggunakan algoritma klasifikasi K-NN cukup baik yaitu 86,55% pada nilai k = 3, serta waktu pemrosesan klasifikasi rata-rata 0,01021 detik per-file musik.DOI: 10.24843/MITE.1601.03
Sistem Klasifikasi Musik Gamelan Angklung Bali Terhadap Suasana Hati Menggunakan Algoritma K-Nearest Neighbor Berbasis Algoritma Genetika
Tria Hikmah Fratiwi;
Made Sudarma;
Nyoman Pramaita
Jurnal Teknologi Elektro Vol 20 No 2 (2021): (Juli-Desember) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.24843/MITE.2021.v20i02.P10
Balinese gamelan angklung instrumental music through its sound waves can interfere the waves of human’s thought to decrease brain’s wave frequency. The aim is to affect the psychological condition in term of mood so it will lead to positive stress level with either low or high energy. Music with positive stress level and low energy level is categorized as contentment. Positive stress level music and high energy level is categorized as exuberance. MIR (Music Information Retrieval) is a part of Data Mining which digging information about music’s data. One of them is the mood’s classification which is interpreted by chunks of music data. This research designs and builds the classification system to detect the Balinese gamelan angklung instrumental music’s mood using K-NN algorithm and K-NN based Genetic Algorithm. K-NN is able to overcome the classification problem well. However, behind its excellence, the very sensitive k value setting becomes a weakness. Applying Genetic Algorithm on the K-NN classification system optimizes the optimal k value’s determination. Based on the same training and testing data set, K-NN gives 81,08% (k=6) as the highest accuracy percentage, while K-NN based Genetic Algorithm gives 89,19% (k=4) as the highest accuracy percentage.
Pemanfaatan Big Data Media Sosial Dalam Menganalisa Kemenangan Pilkada
Dewa Ayu Putri Wulandari;
Made Sudarma;
Nyoman Paramaita
Jurnal Teknologi Elektro Vol 18 No 1 (2019): (Januari - April) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (175.726 KB)
|
DOI: 10.24843/MITE.2019.v18i01.P15
The election of the Governor and Deputy Governor of Bali will go through several stages of elections starting from the determination of the Governor and Deputy Governor of Bali to the stages of vote counting. In the election of the Governor and Deputy Governor of Bali the community can be directly involved in the voting stage which will be held on June 27, 2018 (General Commission Election or KPU, 2018). So that it raises many opinions, not only positive and neutral opinions but also negative ones. This research is expected to be able to conduct research on public opinion which contains positive, neutral and negative sentiments. In this research used tokenization preprocessing data N-gram method. N-gram is a token consisting of three words in each one token. In the stemming stages used the Nzief Adriani algorithm. For the classification process of this research used the ‘Naïve Bays Classifier (NBC) method. In testing the candidate Governor's data the highest accuracy was obtained from the classification ofKBS-Ace data on data taken from twitter with 89% accuracy, 91% precision and 94% recall and lowest accuracy when KBS-Ace data calcification process on social media Facebook.
Penilaian Tingkat Kepuasan Layanan Produk/Jasa dengan Metode Service Quality Model
I Putu Sugi Almantara;
Made Sudarma;
Ida Bagus Alit Swamardika
Jurnal Teknologi Elektro Vol 20 No 2 (2021): (Juli-Desember) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.24843/MITE.2021.v20i02.P06
Product / service satisfaction is one of several factors that can influence a company. Satisfaction is something that can be measured to improve and achieve goals in a company / government institution. The level of satisfaction can be measured by utilizing several methods, including Service Quality. Service Quality can show the difference between the customer's expectations of the company and the services received by the customer and the difference between the expectations and the services received can be seen the level of satisfaction provided by the customer. This study aims to determine the application of the Service Quality method in measuring the level of product / service satisfaction at companies or government institutions with the concept of literature review. The results of the literature review show that the application of the Service Quality method is more widely used in private companies than government institutions. The results also show that product / service service satisfaction is better at government institutions compared to private companies. This is shown in one of the literature which resulted in the achievement of very good levels of satisfaction with service quality in government institutions.
APLIKASI VERIFIKASI WAJAH UNTUK ABSENSI PADA PLATFORM ANDROID DENGAN MENGGUNAKAN ALGORITMA FISHERFACE
I Putu Putrayana Wardana;
IA Dwi Giriantari;
Made Sudarma
Jurnal Teknologi Elektro Vol 15 No 2 (2016): (July - December) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (955.427 KB)
The face is one part of the human body which are often used in biometric recognition system for high-level differences between the faces of the other faces. Android mobile application with additional security face recognition feature will add to the security of personal privacy of a person's use of telephone / mobile in particular that based on android. Extraction is one of the characteristics of the stages through which the development of biometric facial recognition systems on attendance face recognition applications. This stage aims to extract information from the face image so that it can be used as the unique features of the face in question. In this paper face recognition feature extraction phase is done by using algorithms Fisherface. The image of the face through the training process to the alignment faces and extraction fisherface which is then matched by comparing the value euclidiannya. The trial results in this study resulted in fisherface algorithm does not affect the change in facial expression, distance and lighting after testing two hundred thirty facial image database will still be able to recognize a person's face.
Peramalan Nilai Tukar Rupiah Terhadap Mata Uang Negara Asia Menggunakan Metode Quantum Neural Network
Putu Risanti Iswardani;
Made Sudarma;
Lie Jasa
Jurnal Teknologi Elektro Vol 20 No 1 (2021): (Januari - Juni ) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.24843/MITE.2021.v20i01.P18
Mata uang merupakan sebuah alat pembayaran yang digunakan di seluruh dunia. Setiap mata uang memiliki nilai yang bervariasi pada setiap negara, sehingga negara yang satu memiliki nilai mata uang yang berbeda dengan negara lainnya. Asia merupakan sebuah benua terbesar dan terluas di dunia. Sebagian besar wisatawan yang mengunjungi Indonesia berasal dari Asia. Sehingga pertukaran mata uang terbanyak yaitu mata uang dari negara-negara yang ada di Asia. Mata uang digunakan sebagai alat tukar antar negara yang bernilai disebut nilai tukar. Perbedaan nilai tukar antara satu negara dengan negara lainnya dipengaruhi oleh beberapa faktor. Salah satu faktor yang mempengaruhi perbedaan nilai tukar yaitu faktor inflasi yang ada di setiap negara. Untuk mengatasi hal tersebut, perlu adanya sebuah permalan nilai tukar yang dapat digunakan untuk memprediksi nilai tukar dimasa mendatang. Quantum Neural Network pada penelitian ini digunakan untuk memprediksi nilai tukar Indonesia dengan Beberapa Negara di Asia. Mata uang yang digunakan pada penelitian ini yaitu pertukaran mata uang Indonesia dengan Singapore, Hongkong dan Japan. Hasil yang didapatkan pada penelitian ini yaitu nilai akurasi pada Prediksi Nilai Mata Uang dengan menggunakan Quantum Neural Network sebesar 99.78% pada Singapore Dollar ke Indonesia Rupiah, 99.57% pada Hongkong Dollar ke Indonesia Rupiah, dan 99.60% pada Japan Yen ke Indonesian Rupiah.
UMKM (Usaha Mikro Kecil dan Mene Implementasi Layanan Cloud Computing Software As a Service Pada Usaha Mikro Kecil dan Menengah
Rifky Lana Rahardian;
Linawati Linawati;
Made Sudarma
Jurnal Teknologi Elektro Vol 17 No 3 (2018): (September - Desember) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (667.13 KB)
|
DOI: 10.24843/MITE.2018.v17i03.P10
MSMEs (Micro and Small and Medium Enterprises) have an important role now in the ASEAN economy. 96% of companies in ASEAN are MSMEs. Although MSMEs have shown their role, they still face various obstacles and constraints such as lack of information, weak branding, promotion and infrastructure. The application of information technology will certainly make it easier for MSMEs to strengthen branding, business processing and enable cooperation with other parties to provide added value to MSMEs. The application of cloud computing or cloud computing is considered suitable for MSMEs that have limited resources in the form of capital, human resources (HR) and information technology infrastructure. Responding to the importance of the application of information technology to these MSMEs, the research aims to design and build systems that can facilitate MSMEs by utilizing cloud computing technology with SAAS (software as a service). The system has been tested by testing carried out using the blackbox method which is considered to be running optimally, functionally the system built can produce the output expected by the MSMEe participants. Usability testing to measure the level of satisfaction (user satisfaction) with the usability scale (SUS) system method is also felt necessary so that the system is designed and built according to user needs and satisfaction. With the results of the calculation of SUS 72.25 which means that it belongs to class B or above the average makes SAAS cloud computing services a system that can be accepted by MSMEs participants in supporting their business.
Analisis Rating Sentimen pada Video di Media Sosial Youtube Menggunakan STRUCT-SVM
Kadek Ary Budi Permana;
Made Sudarma;
Wayan Gede Ariastina
Jurnal Teknologi Elektro Vol 18 No 1 (2019): (Januari - April) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (327.817 KB)
|
DOI: 10.24843/MITE.2019.v18i01.P17
Sentiment analysis on comments can be used to determine sentiment rating. The comments used are comments on Youtube. The type of video used is the official trailer video Indonesian movie. This paper contains steps to determine sentiment rating by notice the structure of comments. The structure of comments is needed because not all comments are relevant to the topic. Classes on comments are divided into seven classes including positive films, neutral films, negative films, positive not films, neutral not film, negative not film, and spam / off-topic. Comments that have a positive film or film negative class are used to determine sentiment rating. The number of likes in comments also determines the sentiment rating. Comment classification using STRUCT-SVM. The results of STRUCT-SVM show accuracy of 69.68% for linear kernels and 71.34% for RBF kernels.
PEMODELAN INTEGRASI NEARLY REAL TIME DATA WAREHOUSE DENGAN SERVICE ORIENTED ARCHITECTURE UNTUK MENUNJANG SISTEM INFORMASI RETAIL
I Made Dwi Jendra Sulastra;
Made Sudarma;
I Nyoman Satya Kumara
Jurnal Teknologi Elektro Vol 14 No 2 (2015): (July - December) Majalah Ilmiah Teknologi Elektro
Publisher : Universitas Udayana
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1103.011 KB)
|
DOI: 10.24843/MITE.2015.v14i02p03
Updates the data in the data warehouse is not traditionally done every transaction. Retail information systems require the latest data and can be accessed from anywhere for business analysis needs. Therefore, in this study will be made data warehouse model that is able to produce the information near real time, and can be accessed from anywhere by end users application. Modeling design integration of nearly real time data warehouse (NRTDWH) with a service oriented architecture (SOA) to support the retail information system is done in two stages. In the first stage will be designed modeling NRTDWH using Change Data Capture (CDC) based Transaction Log. In the second stage will be designed modeling NRTDWH integration with SOA-based web service. Tests conducted by a simulation test applications. Test applications used retail information systems, web-based web service client, desktop, and mobile. Results of this study were (1) ETL-based CDC captures changes to the source table and then store it in the database NRTDWH with the help of a scheduler; (2) Middleware web service makes 6 service based on data contained in the database NRTDWH, and each of these services accessible and implemented by the web service client.