Summary

Research on resource efficient machine learning covering both inference and training dimensions. Techniques include sparseness, advanced sparseness, model parallelism, reduced precision for memory efficiency, and data parallelism, optimization algorithms, data reuse improvement for time efficiency.