Convolutional Neural Networks and Pattern Recognition: Application to Image Classification
This research study focuses on pattern recognition using convolutional neural network. Deep neural network has been choosing as the best option for the training process because it produced a high percentage of accuracy. We designed different architectures of convolutional neural network in order to find the one with high accuracy of image classification and optimum bias. We used CIFAR-10 data set that contains 60000 Images to train our model on architectures. The best architecture was able to classify images with 95.55% of accuracy and an error of 0.32% using cross validation method. We note that, the numbers of epoch while running the model and the depth of the architecture are factors that contributed to get this performance.
Keywords: Convolutional Neural Networks, Cross validation, Deep Learning, Overfitting, Deep neural network
Download Full-Text
ABOUT THE AUTHORS
Christy Ntambwe Kabamba
First Author received his B.Sc. in Computer Science degree in 2014 from Université Notre-Dame du Kasayi, Kananga, DR Congo. He is currently pursuing his M.Sc. studies at the Department of Computer Science and Mathematics of Université de Kinshasa, Kinshasa, DR Congo. His research interests include the Machine Learning, Software Engineering and Computer Networking. He works as visiting lecturer at the Université de Kananga (Kananga, DR Congo), Université Reverend KIM and Université Orthodoxe au Congo (Kinshasa, DR Congo) since October 2016.
Lucie Mpuekela
Second Author received his B.Sc. in Computer Science degree from Université de Mbujimayi, Mbujimayi, DR Congo, in 2012. She is currently pursuing his M.Sc. studies at the Department of Computer Science and Mathematics of Université de Kinshasa, Kinshasa, DR Congo. Research area: Data analysis.
Simon Ntumba Badibanga
Third Author is Professor and Head of Mathematics and Computer Science department of the Université de Kinshasa. As publications, Author of many publications, such as: "Enhanced Parallel Skyline on multi-core architecture with lax Memory space Cost", IJCSI, volume 13, Issue 5, September 2016, Data mart approach for stock management model with a calendar under budgetary constraint, IJCSI, volume 15, Issue 5, September 2018, Poster et the 2nd International conference on Big Data Analysis and Data Mining, San Antonio, USA, 30 november-01 December 2015 "; Data Mart Approach for Stock Management Model with a calendar Uner Budgetary constraint, IJCSI, volume 15, Issue 5, September 2016.
Eugene
Fourth author is professor at the Mathematics and Computer Science department of the Université de Kinshasa. Director of the Computer Science laboratory at the faculty of Sciences, at the Université de Kinshasa. He is author of many articles in many scientific journals like in IJCSI .Poster and the 2nd International conference on Big Data Analysis and Data Mining, San Antonio, USA, 30 November-01 December 2015.
Christy Ntambwe Kabamba
First Author received his B.Sc. in Computer Science degree in 2014 from Université Notre-Dame du Kasayi, Kananga, DR Congo. He is currently pursuing his M.Sc. studies at the Department of Computer Science and Mathematics of Université de Kinshasa, Kinshasa, DR Congo. His research interests include the Machine Learning, Software Engineering and Computer Networking. He works as visiting lecturer at the Université de Kananga (Kananga, DR Congo), Université Reverend KIM and Université Orthodoxe au Congo (Kinshasa, DR Congo) since October 2016.
Lucie Mpuekela
Second Author received his B.Sc. in Computer Science degree from Université de Mbujimayi, Mbujimayi, DR Congo, in 2012. She is currently pursuing his M.Sc. studies at the Department of Computer Science and Mathematics of Université de Kinshasa, Kinshasa, DR Congo. Research area: Data analysis.
Simon Ntumba Badibanga
Third Author is Professor and Head of Mathematics and Computer Science department of the Université de Kinshasa. As publications, Author of many publications, such as: "Enhanced Parallel Skyline on multi-core architecture with lax Memory space Cost", IJCSI, volume 13, Issue 5, September 2016, Data mart approach for stock management model with a calendar under budgetary constraint, IJCSI, volume 15, Issue 5, September 2018, Poster et the 2nd International conference on Big Data Analysis and Data Mining, San Antonio, USA, 30 november-01 December 2015 "; Data Mart Approach for Stock Management Model with a calendar Uner Budgetary constraint, IJCSI, volume 15, Issue 5, September 2016.
Eugene
Fourth author is professor at the Mathematics and Computer Science department of the Université de Kinshasa. Director of the Computer Science laboratory at the faculty of Sciences, at the Université de Kinshasa. He is author of many articles in many scientific journals like in IJCSI .Poster and the 2nd International conference on Big Data Analysis and Data Mining, San Antonio, USA, 30 November-01 December 2015.
