Soft Computing
https://doi.org/10.1007/s00500-017-2954-3
METHODOLOGIES AND APPLICATION
Enhanced quantum-based neural network learning and its application
to signature verification
Om Prakash Patel
1
· Aruna Tiwari
1
· Rishabh Chaudhary
1
· Sai Vidyaranya Nuthalapati
1
· Neha Bharill
1
·
Mukesh Prasad
2
· Farookh Khadeer Hussain
2
· Omar Khadeer Hussain
3
© Springer-Verlag GmbH Germany, part of Springer Nature 2017
Abstract
In this paper, an enhanced quantum-based neural network learning algorithm (EQNN-S) which constructs a neural network
architecture using the quantum computing concept is proposed for signature verification. The quantum computing concept is
used to decide the connection weights and threshold of neurons. A boundary threshold parameter is introduced to optimally
determine the neuron threshold. This parameter uses min, max function to decide threshold, which assists efficient learning.
A manually prepared signature dataset is used to test the performance of the proposed algorithm. To uniquely identify the
signature, several novel features are selected such as the number of loops present in the signature, the boundary calculation,
the number of vertical and horizontal dense patches, and the angle measurement. A total of 45 features are extracted from each
signature. The performance of the proposed algorithm is evaluated by rigorous training and testing with these signatures using
partitions of 60–40 and 70–30%, and a tenfold cross-validation. To compare the results derived from the proposed quantum
neural network, the same dataset is tested on support vector machine, multilayer perceptron, back propagation neural network,
and Naive Bayes. The performance of the proposed algorithm is found better when compared with the above methods, and
the results verify the effectiveness of the proposed algorithm.
Keywords Quantum computing · Neural network · Signatures · Feature extraction · Classification
Communicated by V. Loia.
B Om Prakash Patel
phd1301201003@iiti.ac.in
B Mukesh Prasad
mukesh.nctu@gmail.com
Aruna Tiwari
artiwari@iiti.ac.in
Rishabh Chaudhary
ee1200231@iiti.ac.in
Sai Vidyaranya Nuthalapati
ee1200221@iiti.ac.in
Neha Bharill
phd12120103@iiti.ac.in
Farookh Khadeer Hussain
Farookh.Hussain@uts.edu.au
Omar Khadeer Hussain
o.hussain@adfa.edu.au
1 Introduction
Artificial neural network has emerged as one of the most
promising areas of research in recent years. It is used for
pattern recognition, classification, mining, clustering and
many more (Castellani 2017; Sartakhti et al. 2017; Wang and
Wang 2010; Gupta and Singh 2011). Researchers have pre-
sented many learning algorithms, such as perceptron, back
propagation and multilayer perceptron for solving both two-
class and multi-class problems. The performance of a neural
network system varies according to its architecture, hidden
layers, the number of neurons in the hidden layer, connection
weights, and threshold. It is also dependent on the features
1
Department of Computer Science and Engineering, Indian
Institute of Technology Indore, Simrol Indore, Madhya
Pradesh, India
2
Centre for Artificial Intelligence, School of Software, FEIT,
University of Technology Sydney, Sydney, Australia
3
School of Business, University of New South Wales,
Canberra, Australia
123