PDF Publication Title:
Text from PDF Page: 032
[37] Real wireless network with 66 base stations. - This work presents two applications for wireless networks traffic forecasting, the prediction of the moment when a specified Base Station (BS) will saturate (long term prediction) and the prediction of traffic anomalies (short term prediction). - Comparison of MLP and ARIMA. For short term predictions MLP is better, for long term predictions ARIMA is better. - Prediction of continuous values (with MAE, MAPE and SMAPE) which makes them impossible to compare with the results of this thesis work (discrete prediction). [93] Several WLAN traffic traces of: 1) Network Research Laboratory at Tianjin University, 2) the Mobile Computing Group at Stanford University and 3) the intranet traffic on the ACM SIGCOMM'01 conference held in the U.C. San Diego in August 2001 SVM outperforms ARIMA, MLP and fractional ARIMA (FARIMA) for both one-step-ahead and multi-step-ahead predictions. - They show that network traffic prediction using [90] CDR recording of Telecom Italia in the city of Milan - They propose several CNN-RNN architectures as an alternative to ARIMA models, obtaining a forecasting accuracy of 70-80% [91] Big proprietary dataset from China Mobile - They present a hybrid deep learning model for spatiotemporal prediction, with an autoencoder for spatial modeling and LSTMs for temporal modeling. They improve the results obtained with ARIMA. [35] Real wireless network for IEEEE802.11 network. - Short term traffic prediction in a large wireless infrastructure. - Prediction of continuous values (with median absolute and relative prediction errors) which makes them impossible to compare with the results of this thesis work (discrete prediction). - Comparison of several Moving Average and ARIMA variants. The best forecasting performance is obtained with exponentially weighted moving average (EMA). - Study of time-series traffic forecasting algorithms [92] GEANT backbone networks - Application of several RNN architectures (LSTM and GRU) to predict network traffic. Traffic prediction [36] - Time series GSM traces of China Mobile - Seasonal ARIMA model to predict wireless traffic workload. The relative error between forecasting values and actual values are all less than 0.02. [94] Real proprietary network traffic and simulated one - Network traffic volume prediction with an HMM model [87] Proprietary WiMAX traffic traces - Authors present a comparison between ARIMA, MLP and Stationary Wavelet Transform, showing that MLP provides better results for volume of Doctoral Thesis: Novel applications of Machine Learning to NTAP - 30PDF 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)