Hadaiq Rolis Sanabila

Automatic Wayang Ontology Construction using Relation Extraction from Free Text

This paper reports on our work to automatically construct and populate an ontology of wayang (Indonesian shadow puppet) mythology from free text using relation extraction and relation clustering. A reference ontology is used to evaluate the generated ontology. The reference ontology contains concepts and properties within the wayang character domain. We examined the influence of corpus data variations, threshold value variations in the relation clustering process, and the usage of entity pairs or entity pair types during the feature extraction stages. The constructed ontology is examined using three evaluation methods, i.e. cluster purity (CP), instance knowledge (IK), and relation concept (RC). Based on the evaluation results, the proposed method generates the best ontology when using a consolidated corpus, the threshold value in relation clustering is 1, and entity pairs are used during feature extraction.

Visualization and statistical analysis of fuzzy-neuro learning vector quantization based on particle swarm optimization for recognizing mixture odors

An electronic nose system had been developed by using 16 quartz resonator sensitive membranes-basic resonance frequencies 20 MHz as a sensor, and analyzed the measurement data through various neural network as a pattern recognition system. The developed system showed high recognition probability to discriminate various single odors even mixture odor to its high generality properties; however the system still need improvement. In order to improve the performance of the proposed system, development of the sensor and other neural network are being sought. This paper explains the improvement of the capability of that system from the point of neural network system. It has been proved from our previous work that FLVQ (Fuzzy Learning Vector Quantization) which is LVQ (Learning Vector Quantization) together with fuzzy theory shows high recognition capability compared with other neural networks, however FLVQ have a weakness for selecting the best codebook vector that will influence the result of recognition. This problem will be anticipated by adding the PSO (Particle Swarm Optimization) method to select the best codebook vector. Then experiment show that the new recognition system (FLVQ-PSO) has produced higher capability compared to the earlier mentioned system.

Automatic fetal head approximation using Particle Swarm Optimization based Gaussian Elliptical Path

The number of maternal and infant mortality in Indonesia is high due to the lack of fetal monitoring. In addition, the number of Obstetrics and Gynecology is limited and it centralized in urban area. In order to resolve that problem, we have built fetal head monitoring system. In this work we examine Particle Swarm Optimization for Gaussian Elliptical Path in automatic fetal head approximation. Particle Swarm Optimization is employ for optimizing ellipse fitting optimization in Gaussian Elliptical Path. Based on the experiment result, Particle Swarm Optimization Based Gaussian Elliptical Path (DoGEll_PSO) has a minimum error compared to Gaussian Elliptical Path (DoGEll).

Generative oversampling method (GenOMe) for imbalanced data on apnea detection using ECG data

One of machine learning problem that is difficult but important to be addressed is imbalanced data where particular data is recessive while the others are dominant. Most of classifiers performance significantly degraded when dealing with imbalanced data. The major approaches to tackle imbalanced data are cost sensitive learning which modifies the classifier and resampling which modifies the data distribution. In this research, we employed generated oversampling method (GenOMe) that generate new data point with a particular distribution as a constraint. We examine three distribution functions: Beta, Gamma, and Gaussian distribution. We use Logistic Regression, Support Vector Machine (SVM), and Naive Bayes as classifier to assure the robustness of GenOMe. The experimental results shows that GenOMe outperforms classification using original data and classification using SMOTe (Synthetic Minority Oversampling Technique) data.

Feature Selection and Reduction for Batik Image Retrieval

Batik is the fabric which is truly unique to Indonesia. Batik image retrieval is the research area which focuses on image processing and image retrieving based on its characteristics. This study investigated the performance of the feature selection and reduction on the batik retrieval process. The feature employed in this experiment is the combination of four feature extraction methods, which are Gabor filter, log-Gabor filter, GLCM, and LBP. SFFS methods is used to carry out the selection of features, meanwhile, PCA is used to perform the reduction feature. Based on the experiment, PCA can increase the precision about 17%. Meanwhile, SFFS can improve the execution time 1800 times faster.

A survey of whole genome alignment tools and frameworks based on Hadoop's MapReduce

Next generation DNA sequencing (NGS) project that aims to give understandings in various genes seems to boosts innovative breakthrough in whole genome issues. Dealing with genomic data requires large-scale data storage and processing. Big data technology could be the most appropriate solution to gaining useful knowledge from data comprehensively. This study discusses about genome tools and framework that implement MapReduce of Hadoops components in sequence alignment computation. The aim of this discussion is presenting an overview of whole genome alignment software tools and the implementation in big data.

Multi codebook LVQ-based artificial neural network using clustering approach

In this paper we proposed multicodebook LVQ-based artificial neural network classifier using clustering approach. The classifiers are LVQ, LVQ2-1, GLVQ, and FNGLVQ. The clustering algorithm used to build multi codebook is K-Means, IK-Means, and GMM. Experiment result shows that on synthteic dataset multi codebook FNGLVQ using GMM clustering has higest improvement with 19,53% mprovement compared to FNGLVQ. Whereas on bencmark dataset multi codebook LVQ2-1 using K-Means clustering has higest improvement with 5,83% improvement cmpared to LVQ-2.1.