عنوان انگلیسی مقاله:
Soft biometric based keystroke classification using PSO optimized neural network
ترجمه فارسی عنوان مقاله:
طبقه بندی نرم افزاری بیومتریک با استفاده از شبکه عصبی بهینه شده PSO
Sciencedirect - Elsevier - Materials Today: Proceedings, Corrected proof: doi:10:1016/j:matpr:2021:01:733
In this work, variable length login-id and password belonging to the user were analyzed for bringing forth a more secure verification system. Soft biometrics such as age group and gender are estimated from key- stroke dynamics patterns when he/she types a given password or login id on a keyboard. Experiments were carried on GREYC a web-based keystroke dataset by exploiting the features from DWT of keystroke dynamics and provides classification results using PSO optimized neural network. Experiments done using PSO-NN resulted in 94% accuracy which clearly out performs the BPNN and GA-NN classifiers.© 2021 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the Emerging Trends in Materials Science, Technology and Engineering.
Keywords: Soft biometric | Discrete wavelet transform (DWT) | Genetic Algorithm optimized neural network (GA-NN) | Back propagation neural network (BPNN) | Particle Swarm Optimized neural network(PSO-NN)