آی اس آی پیپر

دانلود رایکان مقاله انگلیسی ISI با موضوع سیستم کشف نفوذ مبتنی بر دسته کننده چند دسته ای مخفی Naive Bayes

عنوان فارسی مقاله:

سیستم کشف نفوذ مبتنی بر دسته کننده چند دسته ای مخفی Naive Bayes

عنوان انگلیسی مقاله:

A network intrusion detection system based on a Hidden Naïve Bayes multiclass classifier

دانلود رایگان مقاله ISI با فرمت PDF:

مشاهده توضیحات کامل و خرید ترجمه فارسی با فرمت ورد تایپ شده:

بخشی از مقاله انگلیسی :

2. Related Work

A classifier function takes each instance of a dataset and maps it to a distinct class by prediction. In the intrusion detection case, a binary classifier assigns the network events to either a normal event class or a malicious event class, whereas a multiclass classi fier further assigns the malicious event class to DoS, probe, U2R or R2L classes

Similar to many other data mining techniques, building the optimum classifiers requires two important tasks: the selection of the input feature (attribute) from a potentially large set of possible features in a given dataset and the selection of the model (optimization) based on the selected features (Hofmann & Sick, 2003). Selecting the right features is challenging, but it must be performed to reduce the number of features for the sake of efficient processing speed and to remove the irrelevant, redundant and noisy data for the sake of predictive accuracy (Huan & Lei, 2005). A multiclass classifier G needs to map the feature space with A features into C classes on a dataset D, which consists of {E1, E2,..., Ei,..., Et} instances