Optical chinese character recognition with a hidden Markov model classifier—a novel approach

1990 ◽  
Vol 26 (18) ◽  
pp. 1530 ◽  
Author(s):  
B.-S. Jeng ◽  
M.-W. Chang ◽  
S.-W. Sun ◽  
C.-H. Shih ◽  
T.-M. Wu
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Character Recognition ◽  
Chinese Character ◽  
Hidden Markov ◽  
Chinese Character Recognition ◽  
Novel Approach

Detection of Mobile Phone Fraud Using Possibilistic Fuzzy C-Means Clustering and Hidden Markov Model

2016 ◽  
Vol 7 (2) ◽  
pp. 23-44 ◽  
Author(s):  
Sharmila Subudhi ◽  
Suvasini Panigrahi ◽  
Tanmay Kumar Behera
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Mobile Phone ◽  
Hidden Markov ◽  
Clustering Methods ◽  
Fuzzy C Means ◽  
Novel Approach ◽  
Incoming Call ◽  
Fuzzy C Means Clustering ◽  
Reality Mining

This paper presents a novel approach for fraud detection in mobile phone networks by using a combination of Possibilistic Fuzzy C-Means clustering and Hidden Markov Model (HMM). The clustering technique is first applied on two calling features extracted from the past call records of a subscriber generating a behavioral profile for the user. The HMM parameters are computed from the profile, which are used to generate some profile sequences for training. The trained HMM model is then applied for detecting fraudulent activities on incoming call sequences. A calling instance is detected as forged when the new sequence is not accepted by the trained model with sufficiently high probability. The efficacy of the proposed system is demonstrated by extensive experiments carried out with Reality Mining dataset. Furthermore, the comparative analysis performed with other clustering methods and another approach recently proposed in the literature justifies the effectiveness of the proposed algorithm.

Exploring More Representative States of Hidden Markov Model in Optical Character Recognition: A Clustering-Based Model Pre-Training Approach

2015 ◽  
Vol 29 (03) ◽  
pp. 1550014 ◽  
Author(s):  
Zhiwei Jiang ◽  
Xiaoqing Ding ◽  
Liangrui Peng ◽  
Changsong Liu
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Character Recognition ◽  
Optical Character Recognition ◽  
Hidden Markov ◽  
Recognition Performance ◽  
Optimization Methods ◽  
Model Parameters ◽  
Model Structure ◽  
Training Approach

Hidden Markov Model (HMM) is an effective method to describe sequential signals in many applications. As to model estimation issue, common training algorithm only focuses on the optimization of model parameters. However, model structure influences system performance as well. Although some structure optimization methods are proposed, they are usually implemented as an independent module before parameter optimization. In this paper, the clustering feature of states in HMM is discussed through comparing the mechanism of Quadratic Discriminant Function (QDF) classifier and HMM. Then, through the clustering effect of Viterbi training and Baum–Welch training, a novel clustering-based model pre-training approach is proposed. It can optimize model parameters and model structure by turns, until the representative states of all models are explored. Finally, the proposed approach is evaluated on two typical OCR applications, printed and handwritten Arabic text line recognition. And it is compared with some other optimization methods. The improvement of character recognition performance proves the proposed approach can make more precise state allocation. And the representative states are benefit to HMM decoding.

Sign in / Sign up

Export Citation Format

Share Document