This project implements three layered artificial neural network using FPGA as accelerator via OpenCL framework.
The ANN classifies 569 breast mass samples into malignant or benign.
The FPGA accelerates the computation by parallelism which is expressed in two ways:
First, parallelizes one sample computation and secondly, compute all samples concurrently.
The ANN was implemented in 5 different versions which demonstrate the trade off between the hardware units usage
of the FPGA and the program run time.
The algorithm run on DE10-Standard board which built around the Intel
System-on-Chip FPGA and combines dual-core Cortex-A9 embedded cores.
