Indonesian Journal of Electrical Engineering and Computer Science Vol. 20, No. 2, November 2020, pp. 1098~1102 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v20.i2.pp1098-1102  1098 Journal homepage: http://ijeecs.iaescore.com The effect of optimizers in fingerprint classification model utilizing deep learning Farah F. Alkhalid Control and Systems Engineering Department, University of Technology, Iraq Article Info ABSTRACT Article history: Received Feb 13, 2020 Revised Apr 16, 2020 Accepted May 1, 2020 Fingerprint is the most popular way to identify persons, it is assumed a unique identity, which enable us to return the record of specific person through his fingerprint, and could be useful in many applications; such as military applications, social appli cations, criminal applications… etc. In this paper, the study of a new model based deep learning is suggested. The focus is directed on how to enhance the training model with the increase of the testing accuracy by applying four scenarios and comparing among them. The effects of two dedicated optimizers are shown and their contrast enhancement is tested. The results prove that the testing accuracy is 85.61% for “Adadelta” optimizer, whereas for “Adam” optimizer, it is 91.73%. Keywords: Adadelta optimizer Adam optimizer CNN Deep learning Histogram equalization Copyright © 2020 Institute of Advanced Engineering and Science. All rights reserved. Corresponding Author: Farah F. Alkhalid, Control and Systems Engineering Department, University of Technology, Iraq. Email: 10352@uotechnology.edu.iq 1. INTRODUCTION The fingerprint is considered one of the most important methods used to identify people, especially it is unique to each person, and cannot be doubled, where people information can be retrieved by relying on the fingerprint. But this print may be latent and not clear, so by using computer vision in many fields of image processing for example improve the intensity of grey image and increase the contrast of image. On the other hand, the propelled deep learning capacities, and especially convolutional neural systems (CNN) specifically, are essentially propelling the best in class in PC vision and example acknowledgment. The deep CNN is an organically propelled variation of multilayer perceptron and speaks to a common deep learning engineering. Deep learning is dynamically showing its huge in areas of valuable application [1]. In this way, researchers are inquiring about and creating Deep learning strategies that are turning out to be progressively ideal. Wong et. al [2, 3] proposed a multi task model to enhance latent fingerprint by increasing the contrast and denoise it based CNN with very satisfied results, Imane Hachchane et. al [4] studied the face detection using fisher vector and bag of visual words with those same CNN features with satisfied results, but didn’t focus on the optimizers and their effects. In [5] proposed a model to identify and recognize fingerprint usin CNN, they focused on the speed of training and presented a very high speed model. In [6] the authors suggested to decrease the quantity of correlations in programmed unique finger impression acknowledgment frameworks with huge databases. The mix of utilizing PC vision calculations in the picture pre-handling level expands the figuring time. In [7] proposed a new study in analyzing the effects of varing the filters on accuracy of CNN model, based classifiers using human face, fingerprint and iris for person identification. In [8] view of examination singularities and edges relating particular focuses. As a result of low- quality pictures, it is extremely hard to get right places of solitary highlights. The creators utilized investigation edge following and bends highlights to characterize fingerprints. An AI calculation that takes a