Embedded Machine Learning

Catalog description

This course will focus on the recent advances in efficient processing of machine learning. Topics include (1) basic machine learning models (inference and training), including deep convolutional neural networks (DCNN), recurrent neural networks (LSTM, GRU, etc.), Transformer (BERT, RoBERTa, DistilBERT, etc.); (2) different applications including object recognition/detection, super resolution, neural machine translation, etc.; (3) effective machine learning accelerations including model compression, quantization, neural architecture search (NAS), GPU and FPGA implementations, dedicated hardware such as Google TPU or IBM TrueNorth; (4) emerging topics, such as federated learning for edge computing systems.