Djalal Er Riyanto
Departemen Ilmu Komputer/Informatika, Fakultas Sains Dan Matematika,Unversitas Diponegoro

Published : 16 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : SAINS DAN MATEMATIKA

Learning Vector Quantization Pada Pengenalan Pola Tandatangan Prabowo, Anindito; Sarwoko, Eko Adi; Riyanto, Djalal Er
JURNAL SAINS DAN MATEMATIKA Volume 14 issue 4 Year 2006
Publisher : JURNAL SAINS DAN MATEMATIKA

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (675.167 KB)

Abstract

ABSTRAK---Pengenalan pola tandatangan dimaksudkan agar komputer dapat mengenali tandatangan dengan cara mengkonversi gambar, baik yang dicetak ataupun ditulis tangan ke dalam kode. Metode yang dipilih dalam pengenalan pola tandatangan ini adalah metode pembelajaran Kohonen Neural Network(Kohonen) dan Learning Vector Quantization(LVQ). Metode Kohonen mengambil bobot awal secara acak, kemudian bobot tersebut di-update hingga dapat mengklasifikasikan diri sejumlah kelas yang diinginkan. Pada metode LVQ bobot awal di-update dengan menggunakan pola yang sudah ada. Dalam penelitian ini, diberikan hasil pengamatan dan perbandingan tentang tingkat keakuratan dan waktu yang dibutuhkan dalam proses pembelajaran terhadap pola tandatangan pada metode Kohonen dan LVQ menggunakan bahasa pemrograman Microsoft Visual Basic 6.0 Enterprise Edition.Kata kunci: metode Kohonen, neural network, metode Learning Vector Quantization
Solving a system of linear equations by QR Factorization Method for Temperature and Altitude Regression Model against Spontaneous-Potential Widowati, Widowati; Setyawan, Agus; Mustafid, Mustafid; Nur, Muhammad; Sudarno, Sudarno; Harmoko, Udi; Adhy, Satriyo; Gunawan, Gunawan; Subagio, Agus; Tjahjana, Heru; Sulpiani, Ririn; Riyanto, Djalal Er; Suhartono, Suhartono; Mukid, Mochammad Abdul; Suseno, Jatmiko Endro
JURNAL SAINS DAN MATEMATIKA Volume 22 Issue 3 Year 2014
Publisher : JURNAL SAINS DAN MATEMATIKA

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (778.71 KB)

Abstract

Many real problems can be represented in the form of multiple linear regression equation. One of those is the relationship between the variables of temperature and altitude of the spontaneous-potential. In order to determine the parameters of the regression equation, the least squares method was used. From here, there was obtained the system of linear equations. In this paper, to solve systems of linear equations, the exact method was used as the exact solution is certainly better than the approached solution. The method used was the QR factorization method. At the QR factorization, the system of linear equations was written in form of matrix equation. Then, the coefficient matrix which the number of rows is m and number of columns is n with linearly independent columns was factored into the matrix Q which has the same size with the matrix A, with orthonormal columns and matrix R was upper triangular. Furthermore, by backward substitution, it could be obtained the exact solution of linear equation system. As verification of this proposed method, a case study was given using data of temperature, altitude, and spontaneous-potential in the geothermal manifestations area, Gedongsongo, Mount Ungaran Semarang. From here, it was obtained the parameters of exact multiple linear regression model which states the relationship between temperature and altitude toward the spontaneous-potential.