Increasing Prediction Accuracy for Human Activity Recognition Using Optimized Hyperparameters.
Niyati R. DarjiSamuel A. AjilaPublished in: IEEE BigData (2020)
Keyphrases
- prediction accuracy
- human activities
- hyperparameters
- model selection
- cross validation
- activity recognition
- bayesian inference
- closed form
- random sampling
- support vector
- improve the prediction accuracy
- bayesian framework
- em algorithm
- maximum a posteriori
- noise level
- prior information
- sensor data
- predictive power
- human computer interaction
- ensemble methods
- maximum likelihood
- action recognition
- sample size
- incremental learning
- markov random field
- parameter space
- missing values
- incomplete data
- noise reduction
- sensor networks
- computer vision
- generative model