DINAR: Enabling Distribution Agnostic Noise Injection in Machine Learning Hardware.
Karthik GanesanViktor KaryofyllisJulianne AttaiAhmed HamodaNatalie Enright JergerPublished in: HASP@MICRO (2023)
Keyphrases
- machine learning
- low cost
- noise level
- hardware and software
- machine learning methods
- machine learning algorithms
- real time
- pattern recognition
- random noise
- gaussian distribution
- decision trees
- data distribution
- information extraction
- additive noise
- noisy data
- explanation based learning
- learning algorithm
- computer vision
- noisy images
- massively parallel
- noise model
- natural language processing
- spatial coordinates
- noise free
- signal to noise ratio
- spatial distribution
- hardware implementation
- training data
- learning tasks
- learning systems
- computational intelligence
- text mining
- multiscale