No abstract

Numeral recognition remains one of the most important problems in pattern recognition. To the best of our knowledge, little work has been done in Devnagari script compared with those for non Indian scripts like Latin, Chinese and Japanese. In this paper we propose an effective method for recognition of isolated Marathi handwritten numerals written in Devnagari script. Fourier Descriptors that describe the shape of Marathi handwritten numerals are used as feature. 64 dimensional Fourier Descriptors represents the shape of numerals, invariant to rotation, scale and translation. Three different classifiers, namely, nearest neighborhood (NN), K-nearest neighborhood (KNN) and Support Vector Machine (SVM) are used independently in order to recognize test numeral. These classifiers are trained with 64 dimensional Fourier Descriptors (FD) of training samples. The proposed system is experimented with a database of 13000 samples of Marathi handwritten numerals using fivefold cross validation method for result computation. An overall recognition rate of 97.05%, 97.04%and 97.85% are obtained for NN, KNN and SVM respectively.

No abstract

The development of wireless mobile ad hoc networks offers the promise of flexibility, low cost solution for the area where there is difficulties for infrastructure network. A key attraction of this mode of communication is their ease of deployment and operation. However, having a good and robust mobile ad hoc networking will depend entirely on security mechanism system in place. Traditional security mechanisms know as firewalls were used for defensive approach to oppose security obstacle. However, firewalls do not fully or completely defeat intrusions. To cope with this limitation, various intrusions detection systems (IDSs) have been proposed to detect such network intrusion activities. The problem encounter for this particular technique of instruction detections technique is that during network monitoring for data collection for anomaly detection, data that does not contribute to detection must be deleted before detection can be processed or application of learning algorithm for detection of abnormal attacks. In this paper we present a novel feature technique for feature selection before learning technique should be applied. The method has been applied into our own data set, and for the detection purpose we have used most of the well reputed three Machine Learning classifiers with the new selected features for performance evaluation and the experiment shows that higher accuracy results could be achieved with only all the 9 features extracted with our own algorithm with the data set created by using RandomForest classifier.

No abstract