CIDR: A Cost-Effective In-Line Data Reduction System for Terabit-Per-Second Scale SSD Arrays.
Mohammadamin AjdariPyeongsu ParkJoonsung KimDongup KwonJangwoo KimPublished in: HPCA (2019)
Keyphrases
- cost effective
- data reduction
- data compression
- low cost
- cost effectiveness
- preprocessing
- instance selection
- data analysis
- knowledge discovery
- data mining
- classification accuracy
- classification rules
- representative subset
- high dimensionality
- rough set theory
- singular value decomposition
- feature selection
- genetic programming
- clustering algorithm
- neural network
- model selection
- dimensionality reduction
- rough sets
- computer vision
- machine learning