EmptyNN: A neural network based on positive and unlabeled learning to remove cell-free droplets and recover lost cells in scRNA-seq data.
Fangfang YanZhongming ZhaoLukas M. SimonPublished in: Patterns (2021)
Keyphrases
- data sets
- training data
- prior knowledge
- learning algorithm
- data analysis
- image data
- data points
- raw data
- data collection
- unsupervised learning
- active learning
- supervised learning
- background knowledge
- learning models
- probability distribution
- database
- small number
- data processing
- human experts
- learning tasks
- positive data
- high dimensional
- missing data
- high dimensional data
- knowledge discovery
- data structure
- learning process