PDF Publication Title:
Text from PDF Page: 123
LSTM layer with two fully connected layers at the end, provides best detection results, both in terms of accuracy and F1. The inclusion of MaxPooling or additional CNN or LSTM layers does not improve results. Batch normalization between CNN layers and the inclusion of some dropout layers at the end of the network do improve results. For this problem, we have 108 distinct service labels to be detected. This is a multi-class classification problem. There are two possible ways to give results in this case: aggregated and One-vs.-Rest results. For One-vs.-Rest, we focus in a particular class (label) and consider the other classes as a single alternative class, simplifying the problem to a binary classification task for each particular class (one by one). In the case of aggregated results, we try to give a summary result for all classes. There are different alternatives to perform the aggregation (micro, macro, samples, weighted), varying in the way the averaging process is done. The performance metrics in Fig. 7 are aggregated metrics using a weighted average. We have used the weighted average provided by scikit-learn [33], to calculate the aggregated F1, precision, and recall scores. In Fig. 8 we provide the One-vs.-Rest metrics for the classification of the first 15 more frequent labels (results obtained with model CNN+RNN-2a). An important observation in Fig. 8 is that for all labels with a frequency higher than 1% (Fig. 1) we achieve accuracy always higher than 98%, and many cases higher than 99%, and an F1 score higher than 0.96. The macro averaged accuracy for these 15 labels is 99.59% (best value in literature). Fig. 8. Performance metrics (one vs. rest) for the classification of several service labels (15 more frequent) B. Impact of features In Table II we can see the influence of the features employed in the learning process. Fig. 2 Doctoral Thesis: Novel applications of Machine Learning to NTAP - 121PDF Image | Novel applications of Machine Learning to Network Traffic Analysis
PDF Search Title:
Novel applications of Machine Learning to Network Traffic AnalysisOriginal File Name Searched:
456453_1175348.pdfDIY PDF Search: Google It | Yahoo | Bing
Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info
Cruising Review Topics and Articles More Info
Software based on Filemaker for the travel industry More Info
The Burgenstock Resort: Reviews on CruisingReview website... More Info
Resort Reviews: World Class resorts... More Info
The Riffelalp Resort: Reviews on CruisingReview website... More Info
CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com (Standard Web Page)