Garuda - Garba Rujukan Digital

BISECTING DIVISIVE CLUSTERING ALGORITHM BASED ON FOREST GRAPH Wahyu Catur Wibowo, Achmad Maududie ,
PROCEEDING IC-ITECHS 2014 PROCEEDING IC-ITECHS 2014
Publisher : PROCEEDING IC-ITECHS 2014

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

Clustering process aims to group all objects based on their proximity. K-Means is oneof clustering algorithms which is categorized as center-based clustering algorithm. In this algorithm, the value of k is an input parameter that has no prior information. If the k is too small then there is one or more clusters in the result that has a big SSE. To reduce the SSE, this kind of cluster has to be split in to two or more sub-clusters. This paper introduces a new method to split a cluster into two clusters (bisecting) based on minimum forest graph which is called Bisecting Minimum Forest Graph (BMFG). To measure the quality, we used two indicators, i.e. information gain and compactness-separation criterion. The result shows that this method gives a better performance compared with Forgy method. Based on the results, BMFG method produced a maximum purity of each cluster and a maximum consistency index for all the runs. On the other hand, Forgy method achieved the maximum consistency only on well distributed data. For CSC index, BMFG and Forgy methods yielded the similar results. However, for the dataset with relatively not well distributed (noisy), BMFG provided a better index of CSC.

PERBAIKAN INISIALISASI K-MEANS MENGGUNAKAN GRAF HUTAN YANG MINIMUM Maududie, Achmad; Wibowo, Wahyu Catur
Prosiding KOMMIT 2014
Publisher : Prosiding KOMMIT

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

K-Means adalah salah satu algoritma clustering yang sangat popular karena kesederhanaan dan kemampuannya dalam menangani data dengan skala besar. Namun demikian algoritma ini sangat sensitif terhadap centroid awal. Perbedaancentroid awal akan memberikan perbedaan hasil clustering dan apabila centroid awal yang diberikan adalah centroid yang tidak baik maka dapat dipastikan hasil clusteringnya juga tidak baik. Artikel ini memuat sebuah metode baru yang dikembangkan penulis untuk meningkatkan kualitas centroid awal melalui teknik perbaikan k yang didasarkan pada graf hutan yang minimum (minimum forest graf). Hasil percobaan yang telah dilakukan menunjukkan bahwa metode inisialisasi menggunakan graf hutan yang minimum menghasilkan centroid awal yang lebih baik dan konsisten dibandingkan metode Forgy. Disamping itu jumlah perulangan yang harus dilakukan dalam proses clustering dengan menggunakan metode ini adalah lebih sedikit (rerata 3,2) dibandingkan metode Forgy (rerata 6,4).

Extended systematic clustering: Microdata protection by distributing semsitive values Widodo Widodo; Wahyu Catur Wibowo; Eko K. Budiardjo; Harry T. Y. Achsan
Bulletin of Electrical Engineering and Informatics Vol 9, No 4: August 2020
Publisher : Institute of Advanced Engineering and Science

Anonymity data for multiple sensitive attributes in microdata publishing is a growing field at present. This field has several models for anonymizing such as k-anonymity and l-diversity. Generalization and suppression became a common technique in anonymize data. But, the real problem in multiple sensitive attributes is sensitive value distribution. If sensitive values do not distribute evenly to each quasi identifier group, it is potentially revealed to sensitive value holder. This research investigated on how the high-sensitive values are distributed evenly into each group. We proposed a novel method/algorithm for distributing high-sensitive values when it forms groups. This method distributes high-sensitive values evenly and varies high-sensitive values in a group. We called our method as extended systematic clustering since it is an extension of systematic clustering method. Diversity metrics was used for evaluating our method. Experiment result showed our method outperformed systematic clustering with average diversity value 0.9719 while systematic clustering 0.3316.

The Implementation of Minimum Forest Graph for Centroid Updating Process on K-Means Algorithm Achmad Maududie; Wahyu Catur Wibowo
INFORMAL: Informatics Journal Vol 3 No 3 (2018): INFORMAL - Informatics Journal
Publisher : Faculty of Computer Science, University of Jember

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.19184/isj.v3i3.10239

K-Means is a well known algorithms of clusteing. It generates some groups based on degree of similarity. Simplicity of implementation, ease of interpretation, adaptability to sparse data, linear complexity, speed of convergence, and versatile in almost every aspect are noble characteristics of this algorithm. However, this algorithm is very sensitive on defining initial centroids process. Giving a bad initial centroid always produces a bad quality output. Due to this weakness, it is recommended to make some runs with different initial centroids and select the initial centroid that produces cluster with minimum error. However, this procedure is hard to achieve a satisfying result. This paper introduces a new approach to minimize the initial centroid problem of K-Means algorithm. This approach focus on centroid updating stage in K-Means algorithm by applying minimum forest graph to produce better new centroids. Based on gain information and Dunn index values, this approach provided a better result than Forgy method when this approach tested on both well distributed and noisy dataset. Moreover, from the experiments with two dimentional data, the proposed approach produced consisten members of each cluster in every run, where it could not be found in Forgy method.

Intermittent Demand Forecasting Using LSTM With Single and Multiple Aggregation Fityan Azizi; Wahyu Catur Wibowo
Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) Vol 6 No 5 (2022): Oktober 2022
Publisher : Ikatan Ahli Informatika Indonesia (IAII)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29207/resti.v6i5.4435

Intermittent demand data is data with infrequent demand with varying number of demand sizes. Intermittent demand forecasting is useful for providing inventory control decisions. It is very important to produce accurate forecasts. Based on previous research, deep learning models, especially MLP and RNN-based architectures, have not been able to provide better intermittent data forecasting results compared to traditional methods. This research will focus on analyzing the results of intermittent data forecasting using deep learning with several levels of aggregation and a combination of several levels of aggregation. In this research, the LSTM model is implemented into two traditional models that use aggregation techniques and are specifically used for intermittent data forecasting, namely ADIDA and MAPA. The result, based on tests on the six predetermined data, the LSTM model with aggregation and disaggregation is able to provide better test results than the LSTM model without aggregation and disaggregation.

Data Quality Management in Educational Data: A Case Study of Statistics Polytechnic Nori Wilantika; Wahyu Catur Wibowo
Jurnal Sistem Informasi Vol. 15 No. 2 (2019): Jurnal Sistem Informasi (Journal of Information System)
Publisher : Faculty of Computer Science Universitas Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21609/jsi.v15i2.848

Every varsity in Indonesia is responsible for ensuring the completeness, the validity, the accuracy, and the currency of its educational data. The educational data is used for implementing higher-education quality assurance system and formulating policies related to universities and majors in Indonesia. Data quality assessment result indicates that educational data in Statistics Polytechnic did not meet completeness, validity, accuracy, and currency criteria. Data quality management maturity has been measured using Loshin’s Data Quality Maturity Model which result is in level 1 to level 2 of maturity. Only the data quality dimensions component has achieved the expected target. Thus, recommendations have been proposed based on the DAMA-DMBOK framework. The activities needed to be carried out are developing and promoting awareness of data quality; defining data quality requirements; profiling, analyzing, and evaluating data quality; define business rules for data quality, establish, and evaluate the data quality services levels, manage problems related to data quality, design and implement operational procedures for data quality management, and monitor operations and performance of data quality management procedures.

Prediksi Kualitas Air Sungai Menggunakan Metode Pembelajaran Mesin: Studi Kasus Sungai Ciliwung: Prediction of River Water Quality Using Machine Learning Methods: Ciliwung River Case Study MOHAMMAD HAEKAL; WAHYU CATUR WIBOWO
Jurnal Teknologi Lingkungan Vol. 24 No. 2 (2023)
Publisher : BRIN Publishing (Penerbit BRIN)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55981/jtl.2023.795

The Ciliwung River is the largest river in Jakarta area, and its water quality tends to decline. To determine river water quality from time to time, real-monitoring is carried out in real time using Online Monitoring (ONLIMO) technology by installing a multiprobes sensor in the Ciliwung River. Eight parameters were monitored, including pH, Dissolved Oxygen, Nitrate, Turbidity, Total Dissolved Solids, Salinity, Electrical Conductivity, and Temperature. In this study, a data science approach using four Machine Learning models, namely Artificial Neural Networks (ANN), Support Vector Machine (SVM), Random Forest, and Naive Bayes, was used to process monitoring data for one year, from January 1 until December 31, 2018, to predict the Ciliwung River water quality. The AAN results showed that the 5-Hidden Layers model has the highest value of Confusion Matrix (Accuracy, Recall, Precision, Specificity, and F1 Score) compared with all tested Hidden Layer of architecture models. In addition, the Random Forest method has the highest confusion matrix value, followed by the ANN method with 5 Hidden Layers, the Naive Bayes method, and the SVM method. The accuracy value of the first three methods is very high, above 89%. Thus, the first three methods work well to predict the quality of Ciliwung River water. ABSTRAK Sungai Ciliwung merupakan sungai terbesar di wilayah Jakarta, kualitas airnya cenderung menurun. Untuk mengetahui kualitas air sungai dari waktu ke waktu dilakukan pemantauan secara real time menggunakan teknologi Online Monitoring (ONLIMO) dengan memasang multiprobes sensor di Sungai Ciliwung. Delapan parameter yang dipantau meliputi pH, Oksigen Terlarut, Nitrat, Kekeruhan, Total Padatan Terlarut, Salinitas, Konduktivitas Listrik, dan Suhu. Pada penelitian ini, pendekatan sains data menggunakan empat model Machine Learning yaitu Jaringan Syaraf Tiruan (JST), Support Vector Machine (SVM), Random Forest, dan Naive Bayes digunakan untuk mengolah data pemantauan selama satu tahun, dari 1 Januari hingga 31 Desember 2018 untuk memprediksi kualitas air sungai Ciliwung. Hasil metode JST menunjukkan bahwa model dengan 5 Hidden Layer memiliki nilai Confusion Matrix (Accuracy, Recall, Precision, Specificity, dan F1 Score) tertinggi dibandingkan dengan semua model arsitektur Hidden Layer yang diuji. Selain itu, metode Random Forest memiliki nilai Confusion Matrix tertinggi, diikuti metode JST dengan 5 Hidden Layer, metode Naive Bayes, dan terakhir metode SVM. Tingkat akurasi ketiga metode pertama sangat tinggi, yaitu di atas 89%. Dengan demikian, ketiga metode pertama bekerja dengan baik untuk memprediksi kualitas air Sungai Ciliwung.

Adoption of E-Wallets in Timor-Leste: An Extended UTAUT Approach Gusmao, Mazarino Neil Araujo Pires Leite; Hidayanto, Achmad Nizar; Isal, Yugo Kartono; Wibowo, Wahyu Catur
Jurnal Informatika Ekonomi Bisnis Vol. 7, No. 2 (June 2025)
Publisher : SAFE-Network

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37034/infeb.v7i2.1191

This study investigates the factors influencing the adoption of e-wallets in Timor-Leste using an extended Unified Theory of Acceptance and Use of Technology framework. Incorporating context-specific variables digital literacy, trust, inertia, merchant availability, and socialization and campaign the research employs a quantitative approach with data collected from 338 respondents through structured questionnaires. Analysis using Partial Least Squares Structural Equation Modeling reveals that four variables performance expectancy (β = 0.325), digital literacy (β = 0.161), socialization and campaign (β = 0.117), and trust (β = 0.321) significantly influence intention to use e-wallets. Trust emerged as the most influential factor, underscoring the need for secure, transparent systems to encourage adoption. Surprisingly, effort expectancy, social influence, digital infrastructure, merchant availability, and inertia were found to be non-significant. The model explains 77.1% of the variance in intention to use, with a high predictive relevance (Q² = 0.753). These findings suggest that user adoption in low-infrastructure contexts depends more on perceived trust and technological competence than on ease of use or peer influence. The results provide strategic insights for policymakers, service providers, and development actors aiming to promote financial inclusion through digital services in emerging economies like Timor-Leste.

Adoption of E-Wallets in Timor-Leste: An Extended UTAUT Approach Gusmao, Mazarino Neil Araujo Pires Leite; Hidayanto, Achmad Nizar; Isal, Yugo Kartono; Wibowo, Wahyu Catur
Jurnal Informatika Ekonomi Bisnis Vol. 7, No. 2 (June 2025)
Publisher : SAFE-Network

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37034/infeb.v7i2.1191

This study investigates the factors influencing the adoption of e-wallets in Timor-Leste using an extended Unified Theory of Acceptance and Use of Technology framework. Incorporating context-specific variables digital literacy, trust, inertia, merchant availability, and socialization and campaign the research employs a quantitative approach with data collected from 338 respondents through structured questionnaires. Analysis using Partial Least Squares Structural Equation Modeling reveals that four variables performance expectancy (β = 0.325), digital literacy (β = 0.161), socialization and campaign (β = 0.117), and trust (β = 0.321) significantly influence intention to use e-wallets. Trust emerged as the most influential factor, underscoring the need for secure, transparent systems to encourage adoption. Surprisingly, effort expectancy, social influence, digital infrastructure, merchant availability, and inertia were found to be non-significant. The model explains 77.1% of the variance in intention to use, with a high predictive relevance (Q² = 0.753). These findings suggest that user adoption in low-infrastructure contexts depends more on perceived trust and technological competence than on ease of use or peer influence. The results provide strategic insights for policymakers, service providers, and development actors aiming to promote financial inclusion through digital services in emerging economies like Timor-Leste.

Deep Neural Network for Speaker Identification Using Static and Dynamic Prosodic Feature for Spontaneous and Dictated Data Rahman, Arifan; Wibowo, Wahyu Catur
JISIP: Jurnal Ilmu Sosial dan Pendidikan Vol 5, No 4 (2021): JISIP (Jurnal Ilmu Sosial dan Pendidikan)
Publisher : Lembaga Penelitian dan Pendidikan (LPP) Mandala

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58258/jisip.v5i4.2279

We can recognize a person by his voice alone. In principle, the sound has a tone (pitch) that is different for each person. This study aims to measure a Deep Neural Network (DNN) performance with static and dynamic prosodic features. Prosodic is information about sound related to tone, intonation, pressure, duration, and rhythm of a person's pronunciation. The data used is dictated and spontaneous voice data that taken from YouTube. The data used consists of three male voices and one female voice. The data is segmented into various duration, 3 seconds, 5 seconds, and 10 seconds. After the data has been segmented, the static prosodic features with 103 dimensions will be extracted and the dynamic prosodic features with 13 dimensions will be extracted too. Each feature and feature combination will be trained and tested using DNN with a ratio of 90:10. The result shows that the 10 seconds segmented data has higher accuracy than the others. Accuracy of static prosodic features is better than dynamic prosodic features. The average accuracy of DNN for static prosodic features is 87.02%. The average accuracy of DNN for dynamic prosodic features is 72.97%. The average accuracy of DNN for combined static and dynamic prosodic features is 87.72%.

Title

Found 14 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 14 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 14 Documents
Search