Claim Missing Document
Check
Articles

Perbandingan Performa Relational, Document-Oriented dan Graph Database Pada Struktur Data Directed Acyclic Graph Pradana Setialana; Teguh Bharata Adji; Igi Ardiyanto
Jurnal Buana Informatika Vol. 8 No. 2 (2017): Jurnal Buana Informatika Volume 8 Nomor 2 April 2017
Publisher : Universitas Atma Jaya Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24002/jbi.v8i2.1079

Abstract

Abstract.Directed Acyclic Graph (DAG) is a directed graph which is not cyclic and is usually employed in social network and data genealogy. Based on the characteristic of DAG data, a suitable database type should be evaluated and then chosen as a platform. A performance comparison among relational database (PostgreSQL), document-oriented database (MongoDB), and graph database (Neo4j) on a DAG dataset are then conducted to get the appropriate database type. The performance test is done on Node.js running on Windows 10 and uses the dataset that has 3910 nodes in single write synchronous (SWS) and single read (SR). The access performance of PostgreSQL is 0.64ms on SWS and 0.32ms on SR, MongoDB is 0.64ms on SWS and 4.59ms on SR, and Neo4j is 9.92ms on SWS and 8.92ms on SR. Hence, relational database (PostgreSQL) has better performance in the operation of SWS and SR than document-oriented database (MongoDB) and graph database (Neo4j).Keywords: database performance, directed acyclic graph, relational database, document-oriented database, graph database Abstrak.Directed Acyclic Graph (DAG) adalah graf berarah tanpa putaran yang dapat ditemui pada data jejaring sosial dan silsilah keluarga. Setiap jenis database memiliki performa yang berbeda sesuai dengan struktur data yang ditangani. Oleh karena itu perlu diketahui database yang tepat khususnya untuk data DAG. Tujuan penelitian ini adalah membandingkan performa dari relational database (PostgreSQL), document-oriented database (MongoDB) dan graph database (Neo4j) pada data DAG. Metode yang dilakukan adalah mengimplentasi dataset yang memiliki 3910 node dalam operasi single write synchronous (SWS) dan single read (SR) pada setiap database menggunakan Node.js dalam Windows 10. Hasil pengujian performa PostgreSQL dalam operasi SWS sebesar 0.64ms dan SR sebesar 0.32ms, performa MongoDB pada SWS sebesar 0.64ms dan SR sebesar 4.59ms sedangkan performa Neo4j pada operasi SWS sebesar 9.92ms dan SR sebesar 8.92ms. Hasil penelitian menunjukan bahwa relational database (PostgreSQL) memiliki performa terbaik dalam operasi SWS dan SR dibandingkan document-oriented database (MongoDB) dan graph database (Neo4j).Kata Kunci: performa database, directed acyclic graph, relational database, document-oriented database, graph database
METODE KLASIFIKASI DATA MINING DAN TEKNIK SAMPLING SMOTE MENANGANI CLASS IMBALANCE UNTUK SEGMENTASI CUSTOMER PADA INDUSTRI PERBANKAN Hairani Hairani; Noor Akhmad Setiawan; Teguh Bharata Adji
Prosiding SNST Fakultas Teknik Vol 1, No 1 (2016): PROSIDING SEMINAR NASIONAL SAINS DAN TEKNOLOGI 7 2016
Publisher : Prosiding SNST Fakultas Teknik

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (350.379 KB)

Abstract

Class imbalance merupakan sebuah permasalahan yang lazim ditemukan pada dataset, dimana disribusi antara class mayoritas (Negative) dan minoritas (positive) tidak seimbang. Dengan kata lain, class mayoritas memiliki jumlah yang lebih banyak dibandingkan class minoritas. Dengan distribusi yang tidak seimbang, metode pada machine learning cenderung keliru mengklasifikasikan class minoritas. Paper ini mengadopsi pendekatan teknik sampling yaitu Algoritma SMOTE untuk menangani permasalahan class imbalance yang dikombinasikan dengan metode klasifikasi yang lainnya yaitu metode J48, SVM, dan Naive Bayes. Berdasarkan hasil pengujian yang telah dilakukan dengan tools weka menggunakan evaluasi kinerja confusion matrix, menunjukkan bahwa metode J48+SMOTE memiliki tingkat akurasi dan sensitivity paling tinggi yaitu sebesar 0,93% dan 0,93%. Sedangkan metode SVM memiliki nilai specificity  yang paling tinggi sebesar 0.99% dan metode Naive Bayes memiliki waktu komputasi yang paling cepat dibandingkan ketiga metode lainnya sebesar 0.38 seconds. Dengan demikian, metode J48+SMOTE mampu menangani class imbalance pada dataset Bank Direct Marketing pada industri perbankan dibandingkan metode SVM dan Naive Bayes. Kata kunci: Algoritma SMOTE; Class Imbalance; Metode Klasifikasi
MENGGUNAKAN DATA MINING UNTUK SEGMENTASI CUSTOMER PADA BANK UNTUK MENINGKATKAN CUSTOMER RELATIONSHIP MANAGEMENT (CRM) DENGAN METODE KLASIFIKASI (AGORITMA J-48, ZERO-R DAN NAIVE BAYES) Maghfirah Maghfirah; Teguh Bharata Adji; Noor Akhmad Setiawan
Prosiding SNST Fakultas Teknik Vol 1, No 1 (2015): PROSIDING SEMINAR NASIONAL SAINS DAN TEKNOLOGI 6 2015
Publisher : Prosiding SNST Fakultas Teknik

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

Abstract

Paper ini akan membahas mengenai salah satu model strategi pemasaran yaitu Customer Segmentation yang membantu pihak bank untuk membagi pasar menjadi kelompok nasabah yang terbedakan dengan kebutuhan, karakteristik atau tingkah laku yang berbada yang mungkin membutuhkan produk atau bauran pemasaran yang terpisah. Customer Segmentation dapat dilakukan dengan bantuan teknik Data Mining. Oleh karena itu, dalam paper ini akan dilakukan analisis dari dataset yang berasal dari data Bank Marketing dari marketing sebuah Bank di Portugis yang berhubungan dengan berlangganan Deposito Bank dengan menggunakan salah satu dari teknik data mining yaitu teknik Classification dengan membandingkan algoritma Naive Bayes, Rules Zero-R, dan Trees J-48. Dan hasil dari penerapan ketiga algoritma tersebut dalam paper ini menunjukkan bahwa dengan algoritma J-48 memberikan hasil terbaik dengan error rate terkecil, yaitu 5.8769%. Sedangkan jika dilihat dari segi efiesiensi waktu dan hasil klasifikasi, algoritma Zero-R memberikan hasil terbaik (0,03 detik). Selanjutnya dari hasil yang telah diperoleh tersebut diharapkan dapat dihasilkan Customer Segmentation yang sesuai dengan kebutuhan bank yang dapat meningkatkan kualitas servis dan revenue dari bank tersebut. Kata Kunci : Bank Customer Segmentation, Classification, Datamining
SPEECH RECOGNITION OF KV-PATTERNED INDONESIAN SYLLABLE USING MFCC, WAVELET AND HMM Syahroni Hidayat; Risanuri Hidayat; Teguh Bharata Adji
Jurnal Ilmiah Kursor Vol 8 No 2 (2015)
Publisher : Universitas Trunojoyo Madura

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28961/kursor.v8i2.63

Abstract

The Indonesian language is an agglutinative language which has complex suffixes and affixes attached on its root. For this reason there is a high possibility to recognize Indonesian speech based on its syllables. The syllable-based Indonesian speech recognition could reduce the database and recognize new Indonesian vocabularies which evolve as the result of language development. MFCC and WPT daubechies 3rd (DB3) and 7th (DB7) order methods are used in feature extraction process and HMM with Euclidean distance probability is applied for classification. The results shows that the best recognition rateis 75% and 70.8% for MFCC and WPT method respectively, which come from the testing using training data test. Meanwhile, for testing using external data test WPT method excel the MFCC method, where the best recognition rate is 53.1% for WPT and 47% for MFCC. For MFCC the accuracy increased asthe data length and the frame length increased. In WPT, the increase in accuracy is influenced by the length of data, type of the wavelet and decomposition level. It is also found that as the variation of state increased the recognition for both methods decreased.
Narrow Window Feature Extraction for EEG-Motor Imagery Classification using k-NN and Voting Scheme Adi Wijaya; Teguh Bharata Adji; Noor Akhmad Setiawan
Proceeding of the Electrical Engineering Computer Science and Informatics Vol 5: EECSI 2018
Publisher : IAES Indonesia Section

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (368.009 KB) | DOI: 10.11591/eecsi.v5.1665

Abstract

Achieving consistent accuracy still big challenge in EEG based Motor Imagery classification since the nature of EEG signal is non-stationary, intra-subject and inter-subject dependent. To address this problems, we propose the feature extraction scheme employing statistical measurements in narrow window with channel instantiation approach. In this study, k-Nearest Neighbor is used and a voting scheme as final decision where the most detection in certain class will be a winner. In this channel instantiation scheme, where EEG channel become instance or record, seventeen EEG channels with motor related activity is used to reduce from 118 channels. We investigate five narrow windows combination in the proposed methods, i.e.: one, two, three, four and five windows. BCI competition III Dataset IVa is used to evaluate our proposed methods. Experimental results show that one window with all channel and a combination of five windows with reduced channel outperform all prior research with highest accuracy and lowest standard deviation. This results indicate that our proposed methods achieve consistent accuracy and promising for reliable BCI systems.
Impact of Matrix Factorization and Regularization Hyperparameter on a Recommender System for Movies Gess Fathan; Teguh Bharata Adji; Ridi Ferdiana
Proceeding of the Electrical Engineering Computer Science and Informatics Vol 5: EECSI 2018
Publisher : IAES Indonesia Section

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (336.245 KB) | DOI: 10.11591/eecsi.v5.1685

Abstract

Recommendation system is developed to match consumers with product to meet their variety of special needs and tastes in order to enhance user satisfaction and loyalty. The popularity of personalized recommendation system has been increased in recent years and applied in several areas include movies, songs, books, news, friend recommendations on social media, travel products, and other products in general. Collaborative Filtering methods are widely used in recommendation systems. The collaborative filtering method is divided into neighborhood-based and model-based. In this study, we are implementing matrix factorization which is part of model-based that learns latent factor for each user and item and uses them to make rating predictions. The method will be trained using stochastic gradient descent with additional tricks and optimization of regularization hyperparameter. In the end, neighborhood-based collaborative filtering and matrix factorization with different values of regularization hyperparameter will be compared. Our result shows that matrix factorization method with lowest regularization hyperparameter outperformed the other methods in term of RMSE score. In this study, the used functions are available from Graphlab and using Movielens 100k data set for building the recommendation systems.
Penghitungan k-NN pada Adaptive Synthetic-Nominal (ADASYN-N) dan Adaptive Synthetic-kNN (ADASYN-kNN) untuk Data Nominal-Multi Kategori Sri Rahayu; Teguh Bharata Adji; Noor Akhmad Setiawan
Jurnal Otomasi Kontrol dan Instrumentasi Vol 9 No 2 (2017): Jurnal Otomasi Kontrol dan Instrumentasi
Publisher : Pusat Teknologi Instrumentasi dan Otomasi (PTIO) - Institut Teknologi Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5614/joki.2017.9.2.5

Abstract

Pada penelitian ini disajikan tentang contoh proses penghitungan k-NN pada teknik oversampling Adaptive Synthetic-Nominal (ADASYN-N) dan Adaptive Synthetic-kNN (ADASYN-kNN) untuk mengatasi masalah ketidakseimbangan (imbalanced) kelas pada dataset dengan fitur nominal-multi categories. Percobaan penghitungan k-NN menggunakan contoh dataset yang memiliki 10 instances dengan 4 fitur, yang mana masing-masing fiturnya memiliki 3 kategori (multi-categories). Contoh dataset untuk percobaan penghitungan tersebut terdistribusi ke dalam 2 kelas, yaitu kelas A terdapat 3 instances dan kelas B dengan 7 instances. Selanjutnya hasil penghitungan k-NN tersebut diujikan pada sebuah dataset dengan fitur nominal-multi categories yang memiliki distribusi kelas yang tidak seimbang. Kemudian dataset di-oversampling dengan metode ADASYN-N dan ADASYN-kNN, kemudian dilakukan uji klasifikasi menggunakan metode Random Forests. Hasil klasifikasi dibandingkan akurasinya antara dataset asli dan dataset dengan teknik oversampling ADASYN-N serta ADASYN-kNN dan menunjukkan bahwa teknik oversampling ADASYN-N dapat meningkatkan akurasi klasifikasi sebanyak 9,05% dari dataset asli, sedangkan ADASYN-kNN meningkatkan akurasi klasifikasi sebanyak 7,84% dari dataset asli. 
Pengaruh Load Balancing Pada Pemrosesan Paralel untuk Kompresi Video Sudaryanto .; Teguh Bharata Adji; Hanung Adi Nugroho
SENATIK STT Adisutjipto Vol 2 (2016): Peran Teknologi dan Kedirgantaraan Untuk Meningkatkan Daya Saing Bangsa
Publisher : Institut Teknologi Dirgantara Adisutjipto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28989/senatik.v2i0.85

Abstract

Communication and multimedia, especially video data processing require very high resource both computing resources and communication traffic. This requires high-end machines such as servers with high specifications are of course very expensive. This results builds a web based application that implements the concept of parallel processing with load balancing process based CPU Usage to compress video files with FFmpeg software.The results are conditioned compression has half the resolution of the original video data. Based on the test results indicate with load balancing process parallel concepts used, the compression process showed an average speed up value of 8.07% faster than paralle Non load balancing process with 2 compressors, 37.57% with 3 compressors, and 41.24% with 4 compressors. The level of processor efficiency by 28.76% more efficient than paralle Non load balancing process with 2 compressors, 37.57% with 3 compressor, and 41.24%  with 4 compressors. Keywords: pemrosesan paralel, kompresi video, Load Balancing, CPU Usage
Data Benchmark pada Google BigQuery dan Elasticsearch Nisrina Akbar Rizky Putri; Widyawan; Teguh Bharata Adji
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 10 No 3: Agustus 2021
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1334.058 KB) | DOI: 10.22146/jnteti.v10i3.1745

Abstract

Nowadays,the cloud is not only a data storage medium but can be used as a medium for managing or analyzing data. Google offers Google BigQueryas a platform capable of managing and analyzing data,while Elasticsearch itself is a search and analysis engine that can be used to analyze data using Kibana. Using a dataset in the form of tweets crawled through http://netlytic.org/,containing the hashtags #COVID19 and #coronavirus, the data will be analyzed and used to compare its performance with benchmarks. Benchmark is a process used to measure and compare performance against an activity so that the desired level of performance is achieved. Data benchmark is performed on both platforms to generate or determine the workload of the platforms. The result obtained in this study is that Google BigQueryhas superior results, both from the upload container for larger datasets than Elasticsearch and with two query testing models.The query management time on Google BigQueryis also shorter and faster than Elasticsearch. Meanwhile, the visualization results from these two platforms have the same percentage amount.
Aspect Category Classification dengan Pendekatan Machine Learning Menggunakan Dataset Bahasa Indonesia SYAIFULLOH AMIEN PANDEGA PERDANA; Teguh Bharata Aji; Ridi Ferdiana
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 10 No 3: Agustus 2021
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1241.472 KB) | DOI: 10.22146/jnteti.v10i3.1819

Abstract

Customer reviews are opinions on the quality of goods or services that consumers perceive. Customer reviews contain useful information for both consumers and providers of goods or services. The availability of a large number of customer reviews on the websiterequires a framework for extracting sentiment automatically. A customer review often contains many aspects, so the Aspect Based Sentiment Analysis (ABSA) should be used to determine the polarity of each aspect. One of the important tasks in ABSA is Aspect Category Detection. The application of Machine Learning Methods for Aspect Category Detection has been mostly done in the English language domain, but in the Indonesian language domain,there are still a few. This study compares the performance of three machine learning algorithms, namely Naïve Bayes (NB), Support Vector Machine (SVM),and Random Forest (RF),on Indonesian language customer reviews using Term Frequency-Inverse Document Frequency (TF-IDF) as term weighting. The results showthat RFperformsthe best,compared to NB and SVM,in three different domains, namely restaurants, hotels,and e-commerce,with the f1-scoresfor each domainare84.3%, 85.7%, and 89.3%.