Agus Subekti
Balai Pengkajian Teknologi Pertanian Kalimantan Barat Jalan Budi Utomo No. 45, Siantan Hulu, Kotak Pos 6150, Pontianak 78061, Kalimantan Barat, Indonesia

Published : 14 Documents
Articles

Found 14 Documents
Search

A HOS-Based Blind Spectrum Sensing in Noise Uncertainty Subekti, Agus; Sugihartono, Sugihartono; Syambas, Nana Rachmana; Suksmono, Andriyan Bayu
Journal of ICT Research and Applications Vol 9, No 1 (2015)
Publisher : ITB Journal Publisher, LPPM ITB

Show Abstract | Original Source | Check in Google Scholar | Full PDF (144.54 KB)

Abstract

Spectrum sensing for cognitive radio is a challenging task since it has to be able to detect the primary signal at a low signal to noise ratio (SNR). At a low SNR, the variance of noise fluctuates due to noise uncertainty. Detection of the primary signal will be difficult especially for blind spectrum sensing methods that rely on the variance of noise for their threshold setting, such as energy detection. Instead of using the energy difference, we propose a spectrum sensing method based on the distribution difference. When the channel is occupied, the distribution of the received signal, which propagates under a wireless fading channel, will have a non-Gaussian distribution. This will be different from the  Gaussian noise when the channel is vacant. Kurtosis, a higher order statistic (HOS) of  the  4th order,  is used as normality test for the test statistic. We measured the detection rate of the proposed method by performing a simulation of the detection process. Our proposed method’s performance proved superior in detecting a real digital TV signal in noise uncertainty.
ADAPTATION OF FIFTY GENOTYPE GOGO RICE ON THREE ENVIRONMENTAL ULTISOL SOIL ACIDITY Subekti, Agus
Widyariset Vol 14, No 2 (2011): Widyariset
Publisher : LIPI-Press

Show Abstract | Original Source | Check in Google Scholar | Full PDF (326.98 KB)

Abstract

Adaptation genotype upland rice at three environment acidity soil of Ultisol have been done in Toho West Kalimantan. This research consist of three acidity environment: environment of pH 4.5 ( without lime), environmental of pH 5.0 ( lime with dose 1 x Al-dd), and environment of pH 5.5 ( lime with dose 1.5 x Al-dd). Each environment arranged in Randomized Block Designs (RBD) with treatment fifty genotype which repeated twice. Result of research indicated that genotypes which is good to be planted at environment of pH 4.5 ( without lime) are: Gajah Mungkur, IR 53234, PR 36, IR 60080 , S-4325-D-1-2-3-1, BP 1153 C-9-60, and BP 720 C-5-Si-60. There are correlation of phenotypic and genotypic indirectional between absorption of P, panicle length, number of fill grain per panicle, and 1.000 fill grain weight with grain weight per clump.
Perancangan dan Implementasi Mapper dan Demapper untuk DVB-T Suyoto, Suyoto; Subekti, Agus; Lukman, Arif
INKOM Journal Vol 5, No 2 (2011)
Publisher : Pusat Penelitian Informatika - LIPI

Show Abstract | Original Source | Check in Google Scholar | Full PDF (328.607 KB)

Abstract

Pada penelitian ini dilakukan perancangan dan implementasi mapper dan demapper untuk DVB-T (Digital Video Broadcaster-Terrestrial). Mapper digunakan untuk memetakan deretan bit digital kedalam symbol-simbol OFDM yang akan masuk ke IFFT, sedangkan demapper digunakan untuk memetkan simbol-simbol OFDM yang keluar dari FFT ke dalam deretan bit digital. Mapper dan demapper menggunakan konstelasi 16 QAM (Quadrature Amplitude Modulation). 4 bit digunakan untuk memetakan setiap simbol OFDM. Perancangan dilakukan dengan menggunakan ISE 9.2i Xilinx. Hasil dari perancangan diimplementasikan pada virtex-4 development board.
Metoda Adaptive Beamforming untuk Cognitive Radio Subekti, Agus
INKOM Journal Vol 8, No 1 (2014)
Publisher : Pusat Penelitian Informatika - LIPI

Show Abstract | Original Source | Check in Google Scholar | Full PDF (385.786 KB)

Abstract

Salah satu peluang pemanfaatan spektrum secara bersama antara  secondary users dan primary users adalah  dengan memanfaatkan  perbedaan sudut datang sinyal (Angle  of  Arrival-  AoA).  Dengan  aplikasi multi antena, arah berkas dari masing-masing dibentuk dan dapat diatur sehingga terfokus dan tidak saling mengganggu karena memberikan interferensi. Pada tulisan ini diusulkan teknik beamforming di sisi penerima. Arah berkas dari larik dibuat maksimum pada arah datang sinyal dan minimum pada arah referensi. Dengan algoritma LMS (Least Mean Square), pembobot dihitung secara iteratif agar memberikan nilai MSE (Minimum Square Error) dari sinyal keluaran larik dan sinyal referensi yang minimum. Algoritma yang diusulkan selanjutnya dicoba disimulasikan untuk beberapa nilai parameter step size.
ELECTROPOLYMERISATION AND CHARACTERISATION OF DOPED-POLYPYRROLE AS HUMIDITY SENSOR Siswoyo, Siswoyo; Nugroho, Trio F.; Zulfikar, Zulfikar; Subekti, Agus
Indonesian Journal of Chemistry Vol 6, No 2 (2006)
Publisher : Universitas Gadjah Mada

Show Abstract | Original Source | Check in Google Scholar

Abstract

A new type of sensing materials for humidity measurement has been developed based on conducting polymer polypyrrole synthesised from pyrrole by adding some dopant compounds, bromide and chloride, it is prepared by potentiodynamic-electropolymerisation technique. Variation of dopant types and concentration has been carried out in order to investigate the effect of this variation to the change of polymeric conductivity when interacting with water vapour. Polypyrrole-Cl (Ppy-Cl) and polypyrrole-Br (Ppy-Br) exhibit a good principal characteristic as sensor candidate namely responding proportionally to humidly variation ranging 30% - 90% relative humidity. Characterisation test for the sensor candidates has been carried out for evaluating their linearity respond toward humidity, their stability in certain period and their reproducibility in some tests. The results show that Ppy-Cl and Ppy-Br showing good linearity respond with R value in a range of 0.95 - 0.99. Their reproducibility and sensitivity were relatively good, however their respond stability were only last in few days. The stability probably is related to the stability of resulted polymeric structure that very affected by synthesis process and dopant used. It is necessary to extend the use of other dopant materials and changing the synthesis process in order to improve sensor stability. In other hand it is also necessary to characterise other performance characteristic of the sensor namely response time, and interference effect of some volatile chemicals and other gases.   Keywords: polypyrrole, potentiodynamic, electropolymerisation, humidity sensor and conducting polymer
Email Spam Filtering Dengan Algoritma Random Forest Ghani, Muhamad Abdul; Subekti, Agus
IJCIT (Indonesian Journal on Computer and Information Technology) Vol 3, No 2 (2018): IJCIT Nov 2018
Publisher : PPPM AMIK BSI Tasikmalaya

Show Abstract | Original Source | Check in Google Scholar | Full PDF (271.061 KB)

Abstract

AbstrakTeknologi berbasis internet sudah menjadi kebutuhan primer. Berdasarkan hasil survey Badan Pusat Statistik bekerjasam dengan APJII, kegiatan pengiriman dan penerimaan email sudah mengalahkan posisi media sosial dengan mencapai 95.75%. Penggunaan email yang sangat intens dapat menimbulkan dampak positif dan negatif. Karena selain selain sebagai alat komunikasi, pada kenyataannya tidak semua orang menggunakan email  dengan baik dan bahkan ada banyak sekali penyalahgunaan email sehingga berpotensi untuk merugikan orang lain. Email yang disalahgunakan ini biasa dikenal sebagai spam atau junkmail (email sampah) yang mana email tersebut berisikan iklan, penipuan dan bahkan virus. Dalam penelitian ini dilakukan perbandingan beberapa metode klasifikasi data mining diantaranya yaitu Algoritma Naïve Bayes, SVM, J48, dan Random Forest dalam memprediksi spam email dengan tujuan agar algoritma terpilih merupakan yang paling akurat. Dari hasil pengujian menggunakan dengan mengukur kinerja dari keempat algoritma tersebut menggunakan Confusion Matrix dan ROC , diketahui bahwa algoritma Random Forest memiliki nilai accuracy paling tinggi, yaitu 94,22 % dan AUC 0,98 diikuti oleh  algoritma J48 dengan accuracy sebesar 92,70% dan AUC 0,95, SVM dengan nilai accuracy 86,48% dan AUC 0,84 dan terendah yaitu metode naive bayes dengan nilai accuracy sebesar 78,87% dan AUC 0,79.Kata kunci: algoritma naive bayes, email spam, J48, random forest, support vector machine AbstractInternet-based technology has become a primary need. Based on the results of a survey by the Central Bureau of Statistics in cooperation with APJII, email sending and receiving activities have outperformed the position of social media by reaching 95.75%. The use of e-mail that is very intense can have positive and negative impacts. Because other than as a means of communication, in reality not everyone uses email well and there is even a lot of email abuse that has the potential to harm others. This misused email is commonly known as spam or junkmail (junk e-mail) which contains e-mail, fraud and even viruses. In this study a comparison of several data mining classification methods including the Naïve Bayes, SVM, J48, and Random Forest algorithms in predicting spam e-mail with the aim that the selected algorithm is the most accurate. From the test results using measuring the performance of the four algorithms using Confusion Matrix and ROC, it is known that the Random Forest algorithm has the highest accuracy value, which is 94.22% and AUC 0.98 followed by the J48 algorithm with accuracy of 92.70% and AUC 0.95, SVM with 86.48% accuracy value and 0.84 AUC and the lowest is the naive bayes method with accuracy value of 78.87% and AUC 0.79.Keyword: J48, naive bayes algorithm, random forest, spam email, support vector machine
PERMODELAN PREDIKTIF AUTISTIC SPECTRUM DISORDER DENGAN ALGORITMA C.45 Farhan, Oktafian; Subekti, Agus
Jurnal Techno Nusa Mandiri Vol 15 No 2 (2018): TECHNO Periode September 2018
Publisher : PPPM Nusa Mandiri

Show Abstract | Original Source | Check in Google Scholar

Abstract

Autism is a developmental disability experienced throughout the life of a patient with Autistic Spectrum Disorder (ASD). The sooner it ishandled, the more likely the child will return to normal. For this reason, a new method is needed that can help parents to quickly recognize thesymptoms of autism in their children. In a previous study conducted by Fadi Fayez Tabhtah a data set was produced to detect whether a child has autism or not. But the research only produces data sets, he does not examine more in which algorithm is suitable for the data sets that have been produced. The data set attributes have some mising value, which invite a question about the accuracy of data. In this study researchers used the CRISP-DM method and test the accuracy of data sets of previous studies using the C.45 algorithm. Furthermore, the WEKA applicationusing feature selection and influence of the missing value for each attribute and find the most significant. These attributes are then tested withthe C.45 algorithm so that the predictive model of the data set is obtained. The A6 attribute of the decision tree calculation does not appear at all as a branch. A new model is obtained where the A6 attribute is omitted, so that when measured by the C.45 algorithm, a better accuracy value isobtained. The results of the new model were then tested on the new questionnaire data, which produced precise predictions.
Peningkatan Kompetensi Penelitian Dan Pengembangan Serta Optimalisasi Kualitas Kepengawasan Supervisi Manajerial Oleh Pengawas Sekolah Di SMA Negeri 1 Wonoayu Kabupaten Sidoarjo Subekti, Agus
Jurnal Edukasi Gemilang (JEG) Vol 4 No 1 (2019): Vol 4 No 1 (2018)
Publisher : MUSYAWARAH KERJA KEPALA SEKOLAH (MKKS) SMP KOTA MADIUN

Show Abstract | Original Source | Check in Google Scholar | Full PDF (181.572 KB)

Abstract

ABSTRAK : Tujuan yang ingin dicapai dari pelaksanaan Penelitian Tindakan Sekolah ini adalah: 1) Melaksanakan dan membuat laporan upaya peningkatan kompetensi penelitian dan pengembangan oleh guru baik secara mandiri maupun terprogram; 2) Menyusun perangkat pembelajaran untuk satu mata pelajaran lengkap; 3) Menyusun laporan observasi pembelajaran di kelas terhadap beberapa guru dalam upaya mencapai standar kompetensi lulusan; 4) Mengembangkan model penilaian yang secara umum dapat dipandang lebih baik dari apa yang telah dikembangkan di sekolah, baik yang menyangkut mekanisme, prosedur, dan instrumen penilaiannya; 5) Melaksanakan pengkajian terhadap program kepengawasan di Cabang Dinas Pendidikan Provinsi Jawa Timur (SMA Negeri 1 Wonoayu Kabupaten Sidoarjo) berkenaan dengan implementasi 8 standar nasional pendidikan (SI, SKL, Proses, Penilaian, Pengelolaan, Sarpras, Tendik, Pembiayaan).Setelah penulis sebagai pengawas sekolah melaksanakan kegiatan Penelitian Tindakan Sekolah mulai awal sampai akhir pelaksanaan kegiatan, maka kompetensi supervisi akademik penulis sebagai pengawas sekolah dapat meningkat, kompetensi guru dalam mengembangkan perangkat pembelajaran (silabus dan RPP) meningkat, dengan melaksanakan supervisi guru dan kepala sekolah, maka kompetensi supervisi akademik penulis semakin meningkat, meningkatnya kompetensi penulis dalam menyusun perangkat pembelajaran
PEMODELAN PREDIKTIF KONSUMSI ENERGI BANGUNAN GEDUNG KOMERSIAL DENGAN ALGORITMA SUPPORT VECTOR MACHINE Indriyanti, Indriyanti; Subekti, Agus
Jurnal Pilar Nusa Mandiri Vol 14 No 2 (2018): PILAR Periode September 2018
Publisher : PPPM Nusa Mandiri

Show Abstract | Original Source | Check in Google Scholar | Full PDF (1102.847 KB)

Abstract

Konsumsi energi bangunan yang semakin meningkat mendorong para peneliti untuk membangun sebuah model prediksi dengan menerapkan metode machine learning, namun masih belum diketahui model yang paling akurat. Model prediktif untuk konsumsi energi bangunan komersial penting untuk konservasi energi. Dengan menggunakan model yang tepat, kita dapat membuat desain bangunan yang lebih efisien dalam penggunaan energi. Dalam tulisan ini, kami mengusulkan model prediktif berdasarkan metode pembelajaran mesin untuk mendapatkan model terbaik dalam memprediksi total konsumsi energi. Algoritma yang digunakan yaitu SMOreg dan LibSVM dari kelas Support Vector Machine, kemudian untuk evaluasi model berdasarkan nilai Mean Absolute Error dan Root Mean Square Error. Dengan menggunakan dataset publik yang tersedia, kami mengembangkan model berdasarkan pada mesin vektor pendukung untuk regresi. Hasil pengujian kedua algoritma tersebut diketahui bahwa algoritma SMOreg memiliki akurasi lebih baik karena memiliki nilai MAE dan RMSE sebesar 4,70 dan 10,15, sedangkan untuk model LibSVM memiliki nilai MAE dan RMSE sebesar 9,37 dan 14,45. Kami mengusulkan metode berdasarkan algoritma SMOreg karena kinerjanya lebih baik.
PREDIKSI KEKAMBUHAN KANKER PAYUDARA DENGAN ALGORITMA C4.5 Rizqiah, Ai Rita; Subekti, Agus
Jurnal Techno Nusa Mandiri Vol 15 No 2 (2018): TECHNO Periode September 2018
Publisher : PPPM Nusa Mandiri

Show Abstract | Original Source | Check in Google Scholar

Abstract

Breast cancer is known as the fifth cause of death based on WHO data in 2015. The risk of developing breast cancer will increase with age, family medical history, personal medical history, caucasian descent, early menstruation, late menopause and others. This study aims to predict the use of Naïve Bayes and C4.5 data mining algorithms to classify the recurrence of cancer patients based on certain attributes in the breast cancer dataset. The data mining process will help identify the range or value of various attributes of what causes breast cancer. The results of this study indicate that the C4.5 algorithm has an accuracy value of 75.5% better than Naïve Bayes which only has an accuracy value of 72.7%.