The MLIP package: moment tensor potentials with MPI and active learning.
Ivan S. NovikovKonstantin GubaevEvgeny V. PodryabinkinAlexander V. ShapeevPublished in: Mach. Learn. Sci. Technol. (2021)
Keyphrases
- active learning
- high order
- higher order
- zernike moments
- parallel algorithm
- parallel implementation
- message passing
- training set
- semi supervised
- training examples
- selective sampling
- general purpose
- high performance computing
- message passing interface
- parallelization strategy
- shared memory
- imbalanced data classification
- learning algorithm
- machine learning
- supervised learning
- learning strategies
- random sampling
- pairwise
- potential functions
- software package
- diffusion tensor
- feature selection
- experimental design
- learning process
- tensor product
- active learning strategies
- tensor space
- markov random field
- semi supervised learning
- relevance feedback
- parallel computing
- parallel programming
- face recognition
- tensor decomposition
- pool based active learning