SiMul: An Algorithm-Driven Approximate Multiplier Design for Machine Learning.
Zhenhong LiuAmir YazdanbakhshTaejoon ParkHadi EsmaeilzadehNam Sung KimPublished in: IEEE Micro (2018)
Keyphrases
- machine learning
- learning algorithm
- preprocessing
- experimental evaluation
- linear programming
- times faster
- machine learning methods
- probabilistic model
- np hard
- computational complexity
- case study
- optimization algorithm
- machine learning systems
- exact solution
- computer vision
- classification algorithm
- detection algorithm
- expectation maximization
- text classification
- significant improvement
- dynamic programming
- theoretical analysis
- segmentation algorithm
- cost function
- machine learning algorithms
- data sets
- information extraction
- improved algorithm
- hardware implementation