ترجمه مقاله نقش ضروری ارتباطات 6G با چشم انداز صنعت 4.0
- مبلغ: ۸۶,۰۰۰ تومان
ترجمه مقاله پایداری توسعه شهری، تعدیل ساختار صنعتی و کارایی کاربری زمین
- مبلغ: ۹۱,۰۰۰ تومان
Spam continues to inflict increased damage. Varying approaches including Support Vector Machine (SVM) based techniques have been proposed for spam classification. However, SVM training is a computationally intensive process. This paper presents a parallel SVM algorithm for scalable spam filtering. By distributing, processing and optimizing the subsets of the training data across multiple participating nodes, the distributed SVM reduces the training time significantly. Ontology based concepts are also employed to minimize the impact of accuracy degradation when distributing the training data amongst the SVM classifiers.