Skripsi

PENERAPAN RANDOM FOREST UNTUK KLASIFIKASI POLA PENGETIKAN BERBASIS KEYSTROKE DYNAMICS MENGGUNAKAN POLA KESALAHAN DAN KOREKSI

---

Security in authentication systems is a crucial aspect of protecting user data and identity. One approach that can be applied is keystroke dynamics, which analyzes typing patterns as a form of behavioral biometric characteristics. This study aims to develop and evaluate a machine learning based classification model using the Random Forest algorithm to identify user typing patterns. The dataset was obtained through the extraction of keystroke dynamics features, including traditional features (mean dwell time) and non-traditional features (average correction delay, correction burst maximum, retype same character ratio, and error indicator). The preprocessed data were used to train a Random Forest model, with hyperparameter selection performed using Randomized Search Cross-Validation based on a k-fold scheme. The evaluation results show that the best parameter combination achieved a cross-validation accuracy of 0.7187, while testing on the test dataset produced a final accuracy of 0.725. The results indicate that the Random Forest model with cross-validation–based hyperparameter optimization yields optimal performance, and that non-traditional keystroke dynamics features provide insights into typing error and correction patterns, which have the potential to be utilized in the development of behavioral-based authentication systems.

---

Availability

Detail Information

Series Title

-

Call Number

T1897492025

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xiv, 109 hlm.; ilus.; tab.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

005.807

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Keamanan Data--Program Data

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment