Papers by Marinus Ignasius J A W A W U A N Lamabelawa
High Education of Organization Archive Quality: Jurnal Teknologi Informasi, 2018
For numerous purposes, time series data are analyzed to understand phenomena or behaviors of vari... more For numerous purposes, time series data are analyzed to understand phenomena or behaviors of variables, and try to find future value. Interpolation is guessing time series data point between the range of data set. Extrapolation is predict or guessing time series data point from beyond the range of data set. In this study, Newton’s Extrapolation is compared with linear and squared extrapolation. Newton’s Extrapolation making the assumption that the observed trend continues for values of x outside the model range. The robustness of prediction using Root Mean Square Error (RMSE) and Mean Average Percentage Error (MAPE). The results of newton’s interpolation with bottom, middle, and top approaches found the best value are middle approach, namely RMSE 76,01 and MAPE 4,65%. In Newton’s Extrapolation, the error values are consistent at bottom, middle, and top approaches, namely RMSE 541,170 anda MAPE 33,19%. Based on data from the Statistics of Indonesia on the percentage and number of p...
HOAQ Journal STIKOM Uyelindo Kupang, 2019
For numerous purposes, time series data are analyzed to understand phenomena or behaviors of vari... more For numerous purposes, time series data are analyzed to understand phenomena or behaviors of variables, and try to find future value. Interpolation is guessing time series data point between the range of data set. Extrapolation is predict or guessing time series data point from beyond the range of data set.
In this study, Newton’s Extrapolation is compared with linear and squared extrapolation.
Newton’s Extrapolation making the assumption that the observed trend continues for values of x outside the model range.
The robustness of prediction using Root Mean Square Error (RMSE) and Mean Average Percentage Error
(MAPE). The results of newton’s interpolation with bottom, middle, and top approaches found the best value are middle approach, namely RMSE 76,01 and MAPE 4,65%. In Newton’s Extrapolation, the error values are
consistent at bottom, middle, and top approaches, namely RMSE 541,170 and MAPE 33,19%.
Based on data from the Statistics of Indonesia on the percentage and number of poor people in East Nusa Tenggara Province in 2010 -2018 is declining trend pattern. The error value with Linear, Quadratic, and
Newton’s Extrapolation shows the robust value results at linear or trend extrapolation, namely RMSE 157,450 and MAPE 7,93%. These results indicate Newton's extrapolation works well on non-linear data and requires a combination method with soft computing methods such as Fuzzy Systems, AG, or ANN.
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, NTT, 2015
Woven motifs from East Nusa Tenggara as a cultural heritage is currently favored various circles,... more Woven motifs from East Nusa Tenggara as a cultural heritage is currently favored various circles, especially for a fashion model. The Quality of production is determined by the material, origin, motifs, and patterns weaving. Differences of motifs are very interesting to research. This study aims to recognize or identify of pattern Woven motifs from East Nusa Tenggara. The identification process is done by digital image processing techniques through the stages of acquisition, enhancement, and segmentation. Segmentation using feature extraction with edge detection by taking two detection methods that Roberts and Canny edge detection. The results of feature extraction and matcing process can be developed in future studies to Content Based Image Retrieval System (CBIRS).The matching process using euclidean distance results the average obtained Canny method is better than Robertsmethod. By calculating the relative percentages obtained better methods Canny 0.12% compared with the method of Roberts. In the performance measurement value using Precision, Recall and Accuracy seen varied results for the 2nd method. On the threshold of 95%, both methods have the same value, while the 99% threshold Roberts method is better than Canny method. The Value of feature extraction performance is a varied show feature extraction can be improved by other methods based on histogram and Discrete Wavelet Transform.
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, NTT, 2017
According to the Central Agency of Statistics (Badan Pusat Statistik), the number of poor populat... more According to the Central Agency of Statistics (Badan Pusat Statistik), the number of poor population puts the province of East Nusa Tenggara as the 3rd poorest province in Indonesia. The number of poor population in 2016 around 1,149,920 people or decreased 0.914% from 2015. This indicates that the Province of NTT is still struggling to reduce poverty ratings and require a comprehensive strategy to reduce poverty. Scientific and computational studies related to time series data analysis such as predictions are needed to enrich methods and strategies on poverty alleviation for stakeholders in need, especially local government. In this research, doing the calculation of time series data using the interpolation method to analyse and determine a data point based on a set of data that has been known. In testing the approximate accuracy by interpolation methods, measurement of prediction accuracy value is using Mean Square Error (MSE) and Mean Average Percentage Error (MAPE). The calculations performed on 21 time series data, MSE and MAPE values on linear interpolation are better than the quadratic interpolation and newton interpolation. In the calculation of quadratic interpolation with the upper approach, the better accuracy value is 6.25% compared with the bottom approach technique is 6.37%. This is indicated by the decreas of trend poverty in time series data. The value MSE and MAPE of Newton's Interpolation Method is larger than linear interpolation and quadratic interpolation. This is caused by poverty data more linear than non linear.
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, 2016
The ikat weaving motifs of East Nusa Tenggara commonly known through fashion. Quality is determin... more The ikat weaving motifs of East Nusa Tenggara commonly known through fashion. Quality is determined by the material woven motifs, names and places of origin. The ikat weaving motifs quality is determined by the material, the name and place of origin. Motif name as a keyword to determine the origin and quality. The pattern are describing mythical figure like nature, plants, animals, and also abstract motives which shows a deep appreciation the power of nature. The problems are the name of motifs known limited person like weavers and sales centers weaving. In this research, analyzed several feature extraction methods are used to identify the image Feature extraction as basic material to develop a digital image retrieval system or content based image retrieval system.This is so that the name of weaving motifs can be known by many people through a computer-based system. Methods of texture-based feature extraction developed consists of edge detection methods, histogram, and wavelet. The results of each feature extraction methods are identified by calculating the similarity between test images and database images using euclidean distance. Furthermore, the results of retrieval of the three methods used to get the performance of a retrieval system using the values of precision, recall, and accuracy. The results of retrieval system performance testing against 20 test images and 60 databases images practiced by calculating the value of Precision, Recall and Accuracy on the threshold of 90%. The results of the percentage value relative accuracy wavelet method and edge detection is better than the 5.09% histogram method. The test results retrieval system performance against 20 test images and 60 database images practiced by calculating the value of Precision, Recall and Accuracy on the threshold of 90%. The results of the percentage relative accuracy values wavelet method and edge detection is 5.09% better than histogram method.At the threshold 95%,the value accuracy of wavelet method is better than the 20.09% edge detection and wavelet method is better than 20.04% of the histogram. At the threshold of 99%, the value of edge detection performance is 14.35% better than wavelet, and better 42.83% of the histogram. Based on the average results, the results of the performance of a retrieval system, wavelet method is better than edge detection and histogram, where the wavelet result 3.84% better on edge detection, and result 19.20% better against the histogram.. Keywords: The ikat weaving motifs, feature extraction,edge detection,histogram, wavelet
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, 2014
ABSTRAK Tenun ikat Rote-Ndao memiliki ciri dan motif yang khas berdasarkan tekstur, warna, dan ke... more ABSTRAK Tenun ikat Rote-Ndao memiliki ciri dan motif yang khas berdasarkan tekstur, warna, dan ketajaman yang dapat diteliti dengan Pengolahan Citra Digital(PCD). PCD diperlukan untuk mengetahui asal, motif, dan ciri suatu tenun melalui sistem yang lebih efisien. Pertanyaan yang dikaji adalah bagaimana mengingat nama, ciri, dan motif tenun secara permanen dan mudah diretrieve. Jika tersimpan berbasis text maka mudah dilupakan dibandingkan secara visual. Dampaknya adalah ketidaktahuanmotif, asal dan nama motif karena keterbatasan akses terhadap media dan nara sumber. Pada Penelitian ini dikembangkan sistem identifikasi citra tenun yang difokuskan pada analisis tekstur dengan ekstraksi fitur berbasis deteksi tepi. Motif citra tenun didigitalisasi melalui tahap akuisisi, enhancement, segmentasi, dengan deteksi tepi. Tahap akhir ekstraksi fitur adalah identifikasi citra dengan mencocokan citra query dan citra latih yang tersimpan dalam basis data spasial. Hasil penelitian menunjukkan ekstraksi ciri berbasis diskontinuitas dapat digunakan pada citra motif yang memiliki tekstur tepian.Pengujian akurasi deteksi tepi dilakukan dengan menghitung Mean Squared Error(MSE) pada metode Kirsch, Robert, Sobel, dan Prewitt. Nilai MSE terkecil menjadi nilai akurasi terbaik. Berdasarkan pengujian terhadap 20 sampel secara rata-rata nilai MSE Kirsch terbaik yakni 0,469. Sedangkan Pengujian pada metode gradien untuk Robert, Sobel, dan Prewitt menghasilkan MSE terkecil pada metode gradien Prewitt yakni 1,092. Keyword: identifikasi, citra tenun, ekstraksi fitur, deteksi tepi, MSE
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, 2014
The lectures scheduling, known as timetabling problems, in STIKOM Uyelindo currently using the sp... more The lectures scheduling, known as timetabling problems, in STIKOM Uyelindo currently using the spreadsheet
that integrates the schedule of all study programs. The spreadsheet technique has not been fully optimized,
because the scheduling problems are complex. It takes a lot of resources and constraints to produce the optimum
scheduling. The problem of resources is we have to combine the offered subjects, the available lecturers, the
rooms, laboratories, time, and college. The hard constraints related to the relationship between lecturers and
the subjects, classroom, laboratory room, availability of courses, and time availability.
Soft constraints relating to the relationship between variables such as total teaching in 1 week, the total number
of subjects scheduled in a day, the number of courses offered, the total number of credit tought by one lecturer
in a week. This paper discusses how the genetic algorithm can optimize the schedule with the objective of
maximizing the number of classrooms used. Genetic algorithm is one of the very precise algorithm in solving a
complex optimization problems, which so difficult for conventional methods. Designing using genetic algorithm
is expected to be implemented in the next stage of research.
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, 2016
Based on BPS data, the number of poor people in Indonesia, especially East Nusa Tenggara (NTT) is... more Based on BPS data, the number of poor people in Indonesia, especially East Nusa Tenggara (NTT) is ranked 3rd province with the highest poverty which is 20.24%. These data showed that the government should focus downgraded such poverty index zoomed depth and severity of poverty index, as well as increased ability to meet basic needs. This program must be driven in a planned and programmed such as data collection, process, analysis of the results, and the evaluation is complete and accurate. Computer-based system as a reliable data base system is needed so that data can be processed into more precise information with scientific approaches such as data forecasting (forecasting), grouping data (clustering) and data mining (mining). In this research approach is a method to predict the poverty data NTT Province namely statistical methods and methods of soft computing. Statistical methods such as autoregressive (AR), moving average (MA) is not reliable in the model predictions of time series data are not stationary. Soft computing forecasting systems such as fuzzy logic to solve problems ketidakstasioneran reliable time series data. Fuzzy systems are used for the prediction of time series data is known as fuzzy time series. The results show the comparison accuracy of statistical methods and soft computing. Predictive variables were compared is the value of reliability (robustness) with MSE and accuracy (accuracy) with a value of MAPE. Based on testing, the MSE method of fuzzy time series stevenson-porter better 2,58% compared with the classical method and the value of MAPE fuzzy time series stevenson-porter 166% better than the classic method. Based on the comparison method of fuzzy time series and the method of MA generated percentages relative comparison of the MSE Fuzzy Time series 2.91% better MA. MAPE value fuzzy time series 152.47% better than MA ABSTRACT Based on BPS data, the number of poor people in Indonesia, especially East Nusa Tenggara (NTT) is ranked 3rd province with the highest poverty which is 20.24%. These data showed that the government should focus downgraded such poverty index zoomed depth and severity of poverty index, as well as increased ability to meet basic needs. This program must be driven in a planned and programmed such as data collection, process, analysis of the results, and the evaluation is complete and accurate. Computer-based system as a reliable data base system is needed so that data can be processed into more precise information with scientific approaches such as data forecasting (forecasting), grouping data (clustering) and data mining (mining). In this research approach is a method to predict the poverty data NTT Province namely statistical methods and methods of soft computing. Statistical methods such as autoregressive (AR), moving average (MA) is not reliable in the model predictions of time series data are not stationary. Soft computing forecasting systems such as fuzzy logic to solve problems ketidakstasioneran reliable time series data. Fuzzy systems are used for the prediction of time series data is known as fuzzy time series. The results show the comparison accuracy of statistical methods and soft computing. Predictive variables were compared is the value of reliability (robustness) with MSE and accuracy (accuracy) with a value of MAPE. Based on testing, the MSE method of fuzzy time series stevenson-porter better 2,58% compared with the classical method and the value of MAPE fuzzy time series stevenson-porter 166% better than the classic method. Based on the comparison method of fuzzy time series and the method of MA generated percentages relative comparison of the MSE Fuzzy Time series 2.91% better MA. MAPE value fuzzy time series 152.47% better than MA
ABSTRAK Berdasarkan data Badan Pusat Statistik(BPS), Provinsi Nusa Tenggara Timur(NTT) termasuk d... more ABSTRAK Berdasarkan data Badan Pusat Statistik(BPS), Provinsi Nusa Tenggara Timur(NTT) termasuk dalam 5 Provinsi sebagai penghasil ternak sapi terbesar dan menjadi daerah yang didorong menjadi sumber ternak nasional. Rata-rata tingkat pertumbuhan sapi dalam 7 tahun terakhir adalah 8,14%. Program NTT sebagai provinsi ternak harus didukung dengan program swasembada daging, karena merupakan salah satu indikator kemajuan suatu negara. Selain itu, perlu dukungan ilmu lain yang konprehensif seperti analisis populasi sapi yang ilmiah, valid, dan terpercaya yang digunakan untuk memprediksi data masa depan berdasarkan data time series. Sistem prediksi yang handal dan akurat dengan metode statistik, riset operasi, dan komputer akan memberikan informasi kepada pemangku kepentingan untuk mengurangi kesalahan dan memperkecil ketidakpastian pada masa depan. Pada penelitian ini, dilakukan analisis data populasi sapi yang dihimpun secara tahunan atau long term time series. Analisis time series berkaitan dengan pola dan konsistensi data yang tidak tergantung pada waktu, nilai rata-rata, varians, dan kovarian. Metode yang diambil adalah exponential smoothing dan moving averages(MAV). Pada exponential smoothing dilakukan simulasi nilai konstanta pemulusan, sedangkan pada MAV dibandingkan simple MAV dan weight MAV dengan mensimulasikan nilai bobot(weight) untuk mendapatkan nilai kehandalan(robustness) dan nilai akurasi(accuracy). Hasil penelitian adalah membandingkan nilai kehandalan dan keakuratan prediksi dengan ukuran RMSE dan MAPE. Hasil perhitungan exponential smoothing didapat nilai α terbaik = 0.999 yakni RMSE=60.119,054, dan MAPE 5,067 %. Hasil simulasi alpha(α) ditemukan makin besar α akurasi prediksi makin baik. Pada Simple MAV didapat hasil RMSE terbaik pada metode MAV Lag 1 yaitu 61.093,937 dan MAPE terbaik 5,065%. Pada perhitungan prediksi dengan Weight MAV didapat hasil RMSE dan MAPE terbaik pada nilai Weight untuk t-2=0.1 dan t-1=80 dengan RMSE=62.106,394, dan MAPE = 4,990 %. Dari simulasi nilai weight pada t-2 dan t-1 ditemukan bahwa jika nilai weight pada t-2 tetap dan pada t-1 membesar maka akurasi prediksi bertambak baik. Nilai t-1 membesar dan konvergen para nilai t-2 ≥ 80. Sebaliknya jika nilai t-1 tetap dan nilai t-2 membesar maka error prediksi menjadi besar. Perbandingan akurasi prediksi Simple MAV dan Weight MAV, didapat nilai RMSE dari simple MAV lebih baik 1,62% dari Weight MAV. Sedangkan nilai MAPE weight MAV lebih baik 1,51% dari Simple MAV. Pada perhitungan perbandingan akurasi prediksi antara exponential smoothing dan MAV terbaik didapat nilai RMSE dari exponential smoothing lebih baik 3,34% dari Weight MAV. Sedangkan hasil persentase perbandingan relatif nilai MAPE weight MAV lebih baik 1,50% dari exponential smoothing. .
penjelasan tentang pendahuluan riset operasi
Jurnal HOAQ - Teknologi Informasi, STIKOM Uyelindo Kupang, NTT, 2013
Content-Based Image Retrieval System (CBIRS) is a matching process to obtain a number of images b... more Content-Based Image Retrieval System (CBIRS) is a matching process to obtain a number of images based on the input image. In CBIRS are indexed on the basis of low-level features, such as color, texture, and shape that can automatically be derived from the visual content of the images. The sandalwood is a plant whose production is influenced by the age of planting, and that is interesting to use the CBIR method on it. The averages of the age production between 20-40 years. The manual Records and expected life of the planting is used to determine the age of the production. The new
method is to retrieve at the color and texture of the circle formed by the rod.
The research was conducted of a previous study using the Haar wavelet level 1. In this research is developed feature extraction method with Haar wavelet level 2, and new approaches to the Daubechies wavelet. the performance of wavelet to image extraction in stem loop sandalwood tested by calculating the performance of the CBIRS which consist of precision, recall, and accuracy.
The result of the retrieval performance with the euclidean distance of 48 query images and 98 databases images. The Haar level 1 and 2 results the average value of 83,80% and 82,91%. The Daubechies level 1 and 2 results an average value of 85,93% and 83,28%. The best retrieval system performance is Haar level 2, the best precision value is 0,92, recall value is 0,47, and accuracy value is 0,69.
Keywords : sandalwoods, CBIRS, Daubechies wavelet, Haar wavelet, euclidean distance.
Conference Presentations by Marinus Ignasius J A W A W U A N Lamabelawa
SEMMAU STIKOM UYELINDO KUPANG NTT, 2019
One of the regional development priorities is to develop NTT as one of the gates and centers of n... more One of the regional development priorities is to develop NTT as one of the gates and centers of national tourism development (Ring of Beauty). Based on BPS data seen an increase in tourists, especially foreign tourist arrivals in the last 3 years namely in 2017 the number of visits 93,455 up 29.91% from 2016., in 2018 the number of visits was 128,241, up 27.13% from 2017. The focus of local government policy NTT, which places tourism as the prime mover (prime mover), provides a positive trend for the tourism climate which leads to improving people's welfare. This is certainly supported by academics both in the field of tourism and competencies related to data analysis and forecasting. In this research approach a comparative analysis of predictive performance is performed, namely reliability (robust) and accuracy of several prediction models such as exponential smoothing (ES), trend analysis Autoregressive (AR) Moving Averages (MA), and variants of ARMA and ARIMA based on time data series of foreign tourist visits. The amount of performance is done by calculating the robustness value of the Root Mean Square Error (RMSE) and accuracy value with the value of Mean Average Percentage Error (MAPE). The results show that time series data patterns tend to be seasonal patterns rather than trend or exponential data patterns. This is indicated by the predictive performance level of Simple MA (SMA) and Weight MA (WMA), better than Exponential Smoothing (ES) and AutoRegressive (AR). The calculations show WMA Lag 3 is more reliable and accurate than SMA, with more RMSE results better 19.36% and MAPE better 23.27%. In addition, WMA Lag 3 is better than AR (1), where RMSE is 2.5% better and MAPE is 74.80% better. In the exponential pattern analysis, it is seen that ES is not good compared to WMA, where RMSE WMA is 23.52% and MAPE is 78.20% better than ES.
Uploads
Papers by Marinus Ignasius J A W A W U A N Lamabelawa
In this study, Newton’s Extrapolation is compared with linear and squared extrapolation.
Newton’s Extrapolation making the assumption that the observed trend continues for values of x outside the model range.
The robustness of prediction using Root Mean Square Error (RMSE) and Mean Average Percentage Error
(MAPE). The results of newton’s interpolation with bottom, middle, and top approaches found the best value are middle approach, namely RMSE 76,01 and MAPE 4,65%. In Newton’s Extrapolation, the error values are
consistent at bottom, middle, and top approaches, namely RMSE 541,170 and MAPE 33,19%.
Based on data from the Statistics of Indonesia on the percentage and number of poor people in East Nusa Tenggara Province in 2010 -2018 is declining trend pattern. The error value with Linear, Quadratic, and
Newton’s Extrapolation shows the robust value results at linear or trend extrapolation, namely RMSE 157,450 and MAPE 7,93%. These results indicate Newton's extrapolation works well on non-linear data and requires a combination method with soft computing methods such as Fuzzy Systems, AG, or ANN.
that integrates the schedule of all study programs. The spreadsheet technique has not been fully optimized,
because the scheduling problems are complex. It takes a lot of resources and constraints to produce the optimum
scheduling. The problem of resources is we have to combine the offered subjects, the available lecturers, the
rooms, laboratories, time, and college. The hard constraints related to the relationship between lecturers and
the subjects, classroom, laboratory room, availability of courses, and time availability.
Soft constraints relating to the relationship between variables such as total teaching in 1 week, the total number
of subjects scheduled in a day, the number of courses offered, the total number of credit tought by one lecturer
in a week. This paper discusses how the genetic algorithm can optimize the schedule with the objective of
maximizing the number of classrooms used. Genetic algorithm is one of the very precise algorithm in solving a
complex optimization problems, which so difficult for conventional methods. Designing using genetic algorithm
is expected to be implemented in the next stage of research.
method is to retrieve at the color and texture of the circle formed by the rod.
The research was conducted of a previous study using the Haar wavelet level 1. In this research is developed feature extraction method with Haar wavelet level 2, and new approaches to the Daubechies wavelet. the performance of wavelet to image extraction in stem loop sandalwood tested by calculating the performance of the CBIRS which consist of precision, recall, and accuracy.
The result of the retrieval performance with the euclidean distance of 48 query images and 98 databases images. The Haar level 1 and 2 results the average value of 83,80% and 82,91%. The Daubechies level 1 and 2 results an average value of 85,93% and 83,28%. The best retrieval system performance is Haar level 2, the best precision value is 0,92, recall value is 0,47, and accuracy value is 0,69.
Keywords : sandalwoods, CBIRS, Daubechies wavelet, Haar wavelet, euclidean distance.
Conference Presentations by Marinus Ignasius J A W A W U A N Lamabelawa
In this study, Newton’s Extrapolation is compared with linear and squared extrapolation.
Newton’s Extrapolation making the assumption that the observed trend continues for values of x outside the model range.
The robustness of prediction using Root Mean Square Error (RMSE) and Mean Average Percentage Error
(MAPE). The results of newton’s interpolation with bottom, middle, and top approaches found the best value are middle approach, namely RMSE 76,01 and MAPE 4,65%. In Newton’s Extrapolation, the error values are
consistent at bottom, middle, and top approaches, namely RMSE 541,170 and MAPE 33,19%.
Based on data from the Statistics of Indonesia on the percentage and number of poor people in East Nusa Tenggara Province in 2010 -2018 is declining trend pattern. The error value with Linear, Quadratic, and
Newton’s Extrapolation shows the robust value results at linear or trend extrapolation, namely RMSE 157,450 and MAPE 7,93%. These results indicate Newton's extrapolation works well on non-linear data and requires a combination method with soft computing methods such as Fuzzy Systems, AG, or ANN.
that integrates the schedule of all study programs. The spreadsheet technique has not been fully optimized,
because the scheduling problems are complex. It takes a lot of resources and constraints to produce the optimum
scheduling. The problem of resources is we have to combine the offered subjects, the available lecturers, the
rooms, laboratories, time, and college. The hard constraints related to the relationship between lecturers and
the subjects, classroom, laboratory room, availability of courses, and time availability.
Soft constraints relating to the relationship between variables such as total teaching in 1 week, the total number
of subjects scheduled in a day, the number of courses offered, the total number of credit tought by one lecturer
in a week. This paper discusses how the genetic algorithm can optimize the schedule with the objective of
maximizing the number of classrooms used. Genetic algorithm is one of the very precise algorithm in solving a
complex optimization problems, which so difficult for conventional methods. Designing using genetic algorithm
is expected to be implemented in the next stage of research.
method is to retrieve at the color and texture of the circle formed by the rod.
The research was conducted of a previous study using the Haar wavelet level 1. In this research is developed feature extraction method with Haar wavelet level 2, and new approaches to the Daubechies wavelet. the performance of wavelet to image extraction in stem loop sandalwood tested by calculating the performance of the CBIRS which consist of precision, recall, and accuracy.
The result of the retrieval performance with the euclidean distance of 48 query images and 98 databases images. The Haar level 1 and 2 results the average value of 83,80% and 82,91%. The Daubechies level 1 and 2 results an average value of 85,93% and 83,28%. The best retrieval system performance is Haar level 2, the best precision value is 0,92, recall value is 0,47, and accuracy value is 0,69.
Keywords : sandalwoods, CBIRS, Daubechies wavelet, Haar wavelet, euclidean distance.