Otentifikasi Tanda-Tangan Menggunakan Algoritma Back Propagantion Lavenberq Marquardt

Authors

  • Yusran Timur Samuel

https://doi.org/10.36342/teika.v4i1.136

Abstract

Abstrak

Sudah sejak lama tanda tangan digunakan untuk proses otentifikasi misalnya pada transaksi perbankan. Pada proses manual biasanya mata manusia yang menentukan apakah suatu tanda tangan asli atau palsu. Namun seharusnya proses tersebut dapat digantikan oleh komputer. Proses tersebut haruslah cepat dan akurat. Penelitian ini mengembangkan satu sistem otentifikasi tanda tangan dengan menggunakan jaringan saraf tiruan algoritma Lavenberq-Marquardt. Agar proses otentifikasi dapat cepat maka proses praprocessing seperti yang dilakukan pada beberapa penelitian harus dihindarkan dengan membuat suatu sistem yang

menggunakan digitizer tablet untuk menuliskan tanda tangan yang hasilnya dapat langsung di proses tanpa melalui tahap praproses.

Pengujian pada penelitian ini menghasilkan rata-rata kecepatan proses otentifikasi adalah 1,16 detik untuk tanda tangan asli dan 1,20 detik untuk tanda tangan palsu dengan tingkat akurasi dari sistem adalah 100%.

 

Abstract

Signature has been used for long time as an authentication process such as in banking transactions. In a manual process, the human eye is usually used to determine whether a signature is genuine or fake. But the process should be replaced by computers. The process must be fast and accurate. This research develops a signature authentication system using artificial neural network with Lavenberq-Marquardt algorithm. In order for the faster authentication process then preprocessing process as is done in some studies should be avoided by creating a system that uses a digitizer tablet to write a signature that results can be

processed directly without going through a phase preprocess.

Testing in this research show that average speed of an authentication process is 1.16 seconds for the original signature and 1.20 seconds for a false signature with the accuracy of the system is 100%.

Article Metrics

Downloads

Download data is not yet available.

References

Abbas, Rasha, 1994, Backpropagation Network Prototype for Signature Verification, Departement of Computer Science, RMIT.

Cyber SIGN, 2012 "Biometric Signature Verification" (hftp://www.cybersign.com/ techoverview-what.htm diakses 1 April 2011).

Demuth, H., Beale, M., 1999, Neural Network Toolbox For Use with MATLAE, The MathWorks, Inc., Natick, MA.

Izzudin, M., 2001, Otentifikasi Tanda Tangan Algoritma Backpropagation, Program Studi Teknik Elektro, Program Pasca Sarjana, ITB.

Mighell, D.A. Wilkinson, T.S. &Goodman, J.W, 1989, Backpropagation and Its Application to Handwritten Signature Verification, Advances in Neural lnformation Processing Systems /, Hal. 340-347, Kaufman, San Mateo, CA.

National Science and Technology Council "Dynamic Signature" (http://www.biometrics.gov/Documents/DynamicSig.pdf, diakses 1 Mei 2011)

Rahmat, Setiawan, R., Purnomo, M. H., 2006, Backpropagation pada Proses Learning Jaringan Saraf Tiruan untuk Pengenalan Pola Sinyal Elektrokardiograf , Seminar Nasional Aplikasi Teknologi Informasi, Hal 39-44.

Sabourin, R., & Drouhard, J.,7992, Signature Verification Using Directional PDF and Neural Networks, Proceeding of 11'n IAPR lnternational Conference on Pattern Recognition, vol. tt conference B: Pattern Recongniton Methodology and Sysfems, Hal. 321-325, IEEE Computer Science Press.

Shen, W., Surette, M., Khanna, R., 1997, Evalution of automated biometrics base identification and verification systems, Proceeding of the IEEE, vol 85, no 9, September 1997, Woodward, )., 1997, Biometrics. Privacy's foe or privacy's friend?, Proceedings of the IEEE, vol 85, no.9, September 1997.

Downloads

Published

2011-06-01

How to Cite

Samuel, Y. T. (2011). Otentifikasi Tanda-Tangan Menggunakan Algoritma Back Propagantion Lavenberq Marquardt. TeIKa, 4(1), 1-8. https://doi.org/10.36342/teika.v4i1.136

Issue

Section

Programming

Most read articles by the same author(s)