This is the website hosting the code and data in the submitted paper,"HardGBM".
This project contains the following directories:
- GBM/ directory containing open source datasets from UCI, Kaggle and data.world, our proposed ensemble learning method, and the generated Verilog code for hardware simulation experiments.
- CMP_CNN_DNN/ directory used for the development of comparative experiments,, including training neural networks using python, and describing the neural network in C++ code used to build Vivado HLS.
- OS: Ubuntu 18.04.5 LTS
- Python 3.8.12
- Vivado 2018.3
- Vivado HLS 2018.3
All hardware experiments are implemented on Xilinx Zynq-7000 FPGA.