Handwritten Digit Recognition Using SVM Binary Classifiers and Unbalanced Decision Trees

被引:4
作者
Gil, Adriano Mendes [2 ]
Fernandes Costa Filho, Cicero Ferreira [1 ]
Fernandes Costa, Marly Guimaraes [1 ]
机构
[1] Univ Fed Amazonas, Ctr Pesquisa & Desenvolvimento Tecnol Eletron & I, UFAM CETELI, Manaus, Amazonas, Brazil
[2] Inst Nokia Tecnol, Manaus, Amazonas, Brazil
来源
IMAGE ANALYSIS AND RECOGNITION, ICIAR 2014, PT I | 2014年 / 8814卷
关键词
Handwritten digit recognition; MNIST database; Support vector machine; Unbalanced decision tree; Binary classifiers;
D O I
10.1007/978-3-319-11758-4_27
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this work, we use SVM binary classifiers coupled with a binary classifier architecture, an unbalanced decision tree, for handwritten digit recognition. According to input variables, two classifiers were trained and tested. One using digit characteristics and the other using the whole image as input variables. Developed recently, the unbalanced decision tree architecture provides a simple structure for a multiclass classifier using binary classifiers. In this work, using the whole image as input, 100% handwritten digit recognition accuracy was obtained in the MNIST database. These are the best results published in the literature for the MNIST database.
引用
收藏
页码:246 / 255
页数:10
相关论文
共 50 条
[31]   Handwritten Oriya Digit Recognition Using Maximum Common Subgraph Based Similarity Measures [J].
Ghosh, Swarnendu ;
Das, Nibaran ;
Kundu, Mahantapas ;
Nasipuri, Mita .
INFORMATION SYSTEMS DESIGN AND INTELLIGENT APPLICATIONS, VOL 3, INDIA 2016, 2016, 435 :165-173
[32]   Devanagari Handwritten Character Recognition using Transfer Learning with Deep CNN and SVM [J].
Ansari, Mohd Saqib ;
Wasid, Mohammed ;
Rahman, Syed Atiqur .
2022 5TH INTERNATIONAL CONFERENCE ON MULTIMEDIA, SIGNAL PROCESSING AND COMMUNICATION TECHNOLOGIES (IMPACT), 2022,
[33]   An automatic method for arterial pulse waveform recognition using KNN and SVM classifiers [J].
Tânia Pereira ;
Joana S. Paiva ;
Carlos Correia ;
João Cardoso .
Medical & Biological Engineering & Computing, 2016, 54 :1049-1059
[34]   An automatic method for arterial pulse waveform recognition using KNN and SVM classifiers [J].
Pereira, Tania ;
Paiva, Joana S. ;
Correia, Carlos ;
Cardoso, Joao .
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING, 2016, 54 (07) :1049-1059
[35]   Comparison of kNN Classifier and Simple Neural Network in Handwritten Digit Recognition Using MNIST Database [J].
Koman, Wiktoria ;
Malecki, Kuba .
INFORMATION AND SOFTWARE TECHNOLOGIES, ICIST 2023, 2024, 1979 :264-273
[36]   Handwritten digit recognition using neural networks and dynamic zoning with stroke-based descriptors [J].
Alvarez-Leon, David ;
Fernandez-Diaz, Ramon-Angel ;
Sanchez-Gonzalez, Lidia ;
Alija-Perez, Jose-Manuel .
LOGIC JOURNAL OF THE IGPL, 2017, 25 (06) :979-990
[37]   Analysis of PCA and LDA Features for Facial Expression Recognition Using SVM and HMM Classifiers [J].
Varma, Satishkumar ;
Shinde, Megha ;
Chavan, Satishkumar S. .
TECHNO-SOCIETAL 2018: PROCEEDINGS OF THE 2ND INTERNATIONAL CONFERENCE ON ADVANCED TECHNOLOGIES FOR SOCIETAL APPLICATIONS - VOL 1, 2020, :109-119
[38]   Hierarchical Sparse Autoencoder Using Linear Regression-based Features in Clustering for Handwritten Digit Recognition [J].
Phan, Hai T. ;
Duong, An T. ;
Nam Do-Hoang Le ;
Tran, Son T. .
2013 8TH INTERNATIONAL SYMPOSIUM ON IMAGE AND SIGNAL PROCESSING AND ANALYSIS (ISPA), 2013, :183-+
[39]   Using Modified Contour Features and SVM Based Classifier for the Recognition of Persian/Arabic Handwritten Numerals [J].
Alaei, Alireza ;
Pal, Umapada ;
Nagabhushan, P. .
ICAPR 2009: SEVENTH INTERNATIONAL CONFERENCE ON ADVANCES IN PATTERN RECOGNITION, PROCEEDINGS, 2009, :391-394
[40]   Use of wavelet-based two-dimensional scaling moments and structural features in cascade neuro-fuzzy classifiers for handwritten digit recognition [J].
Cetisli, Bayram ;
Edizkan, Rifat .
NEURAL COMPUTING & APPLICATIONS, 2015, 26 (03) :613-624