%0 Journal Article %1 noauthororeditor %A Gyanchandani, Manasi %A J.L.Rana, %D 2014 %J International Journal of Computational Science, Information Technology and Control Engineering (IJCSITCE) %K Anomaly Detection Intrusion System based classificationalgorithm %N 2 %P 37-46 %T Search for Optimized Cost matrix for Performance Enhancement of Anomaly Based Intrusion Detection System using Cost Sensitive Classifier %U https://airccse.com/ijcsitce/papers/1214ijcsitce03.pdf %V 1 %X Intrusion Detection Systems (IDSs) should maximize security while minimize cost. Classic evaluation measures have been extensively studied in the past. Recently, cost-sensitive classification has received much attention. A cost-sensitive classifier uses cost values to evaluate the performance of the classifier. However, these cost values must be given in advance and are generally unknown for a given dataset. It is very time consuming to find these cost values. Again if it is possible to find out such cost values same cannot be used for other datasets. In a typical classification task, all types of misclassifications are treated equally. However, in many practical cases, not all misclassifications are equal. Therefore, it is critical to use a cost-sensitive classifier to minimize cost of misclassifications. This work uses MetaCost, a costsensitive meta-classifier that takes in a classification algorithm, training data, and a cost matrix. In order for MetaCost to be effective, we need to find an optimal cost matrix. In this paper we have proposed a new optimization technique for choosing the cost matrix: cost matrix optimization technique for Anomaly Based Intrusion Detection System (ABIDS). This approach can be applied for finding out optimized cost matrix for any datasets.

@article{noauthororeditor, abstract = {Intrusion Detection Systems (IDSs) should maximize security while minimize cost. Classic evaluation measures have been extensively studied in the past. Recently, cost-sensitive classification has received much attention. A cost-sensitive classifier uses cost values to evaluate the performance of the classifier. However, these cost values must be given in advance and are generally unknown for a given dataset. It is very time consuming to find these cost values. Again if it is possible to find out such cost values same cannot be used for other datasets. In a typical classification task, all types of misclassifications are treated equally. However, in many practical cases, not all misclassifications are equal. Therefore, it is critical to use a cost-sensitive classifier to minimize cost of misclassifications. This work uses MetaCost, a costsensitive meta-classifier that takes in a classification algorithm, training data, and a cost matrix. In order for MetaCost to be effective, we need to find an optimal cost matrix. In this paper we have proposed a new optimization technique for choosing the cost matrix: cost matrix optimization technique for Anomaly Based Intrusion Detection System (ABIDS). This approach can be applied for finding out optimized cost matrix for any datasets.}, added-at = {2018-08-31T14:08:48.000+0200}, author = {Gyanchandani, Manasi and J.L.Rana}, biburl = {https://www.bibsonomy.org/bibtex/24b534d25810ed88ccd1cde3029954f02/ashvath}, interhash = {5ebc9d757125b5123e725185a3639ba7}, intrahash = {4b534d25810ed88ccd1cde3029954f02}, journal = {International Journal of Computational Science, Information Technology and Control Engineering (IJCSITCE) }, keywords = {Anomaly Detection Intrusion System based classificationalgorithm}, month = {july}, number = 2, pages = {37-46}, timestamp = {2018-08-31T14:08:48.000+0200}, title = {Search for Optimized Cost matrix for Performance Enhancement of Anomaly Based Intrusion Detection System using Cost Sensitive Classifier }, url = {https://airccse.com/ijcsitce/papers/1214ijcsitce03.pdf}, volume = 1, year = 2014 }