PDF Publication Title:
Text from PDF Page: 171
Figure 3. Option A. VGM model with Gaussian and Bernoulli distributions. 3.3.2. Option B. Model with Gaussian and Bernoulli distributions plus RMSE for continuous features Option B is presented in Figure 4. This option is similar to option A, but we have divided the output layer according to the separation between discrete and continuous features. We treat the discrete features as in Option A, and for the continuous features we change the loss function to the root mean square error (RMSE) between original and generated continuous features, instead of the log-likelihood, as in option A. We have tried this change in the loss function to see whether we could obtain an improvement by separating the behavior of continuous and discrete features. The change can be also justified by considering that minimizing an RMSE loss function is equivalent to minimizing the negative log-likelihood of an implicit Gaussian distribution for the last decoder layer (in accordance with ELBO theory). In Figure 4, in squared boxes are the elements of the loss function to be minimized for this option. Doctoral Thesis: Novel applications of Machine Learning to NTAP - 169PDF 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)