PDF Publication Title:
Text from PDF Page: 043
Fig 4. Relationship between machine learning, deep learning and generative models Machine learning: Pros • Solid and well-known methods • Used in many applications • Easy interpretability of results for some models (e.g. decision trees) Cons • Requires extensive feature engineering • Many models available but not integrable during the learning phase. Ensemble methods are the way to integrate them, but only at the results level. • Difficult interpretability of results for some models (e.g. random forest) Deep learning: Pros • State of the art results for some problems • Requires less feature engineering • Good for representation learning and to create data embeddings • Framework in which many models can be integrated (in a Lego-way) as long as the resulting loss and activation functions are differentiable. Cons • Generally, requires large datasets • Many hyper-parameters and difficult tuning • High computational resources demanded for training • Lack of interpretability (difficult justification of algorithm decisions) Generative models: Pros • Based on learning the probability distribution of the data (features and labels jointly) • Allows the sampling of the probability distribution to generate new data or imputation of missing features. • Possibility to learn a latent representation of the data simultaneously (e.g. VAE) Doctoral Thesis: Novel applications of Machine Learning to NTAP - 41PDF 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)