resource constrained device
Shallow RNN: Accurate Time-series Classification on Resource Constrained Devices
Recurrent Neural Networks (RNNs) capture long dependencies and context, and 2 hence are the key component of typical sequential data based tasks. However, the sequential nature of RNNs dictates a large inference cost for long sequences even if the hardware supports parallelization. To induce long-term dependencies, and yet admit parallelization, we introduce novel shallow RNNs. In this architecture, the first layer splits the input sequence and runs several independent RNNs.
Shallow RNN: Accurate Time-series Classification on Resource Constrained Devices
Recurrent Neural Networks (RNNs) capture long dependencies and context, and 2 hence are the key component of typical sequential data based tasks. However, the sequential nature of RNNs dictates a large inference cost for long sequences even if the hardware supports parallelization. To induce long-term dependencies, and yet admit parallelization, we introduce novel shallow RNNs. In this architecture, the first layer splits the input sequence and runs several independent RNNs. We provide theoretical justification for our architecture under weak assumptions that we verify on real-world benchmarks.
Shallow RNN: Accurate Time-series Classification on Resource Constrained Devices
Dennis, Don, Acar, Durmus Alp Emre, Mandikal, Vikram, Sadasivan, Vinu Sankar, Saligrama, Venkatesh, Simhadri, Harsha Vardhan, Jain, Prateek
Recurrent Neural Networks (RNNs) capture long dependencies and context, and 2 hence are the key component of typical sequential data based tasks. However, the sequential nature of RNNs dictates a large inference cost for long sequences even if the hardware supports parallelization. To induce long-term dependencies, and yet admit parallelization, we introduce novel shallow RNNs. In this architecture, the first layer splits the input sequence and runs several independent RNNs. We provide theoretical justification for our architecture under weak assumptions that we verify on real-world benchmarks.