Increasing Robustness for Machine Learning Services in Challenging Environments: Limited Resources and No Label Feedback.
Lucas BaierNiklas KühlJörg SchmittPublished in: IntelliSys (1) (2021)
Keyphrases
- limited resources
- machine learning
- computing resources
- processing power
- highly dynamic
- computing environments
- real world
- service oriented
- machine learning methods
- machine learning algorithms
- natural world
- service providers
- heterogeneous environments
- service composition
- reinforcement learning
- web services
- intelligent environments
- active learning
- data mining
- service discovery
- class labels
- artificial intelligence
- information services
- web environment
- resource limitations
- home environments
- multi label
- ubiquitous computing
- semi supervised learning
- end users
- learning algorithm
- dynamic environments
- context aware
- text classification
- text mining
- multi label classification
- natural language processing
- ubiquitous computing environments
- supervised learning
- decision trees
- computer vision
- increasing number of users