TensorGP – Genetic Programming Engine in TensorFlow

In this paper, we resort to the TensorFlow framework investigate benefits of applying data vectorization and fitness caching methods domain evaluation in Genetic Programming. For purpose, an independent engine was developed, TensorGP, along with a testing suite extract comparative timing results across different architectures amongst both iterative vectorized approaches. Our performance benchmarks demonstrate that by exploiting eager execution model, gains up two orders magnitude can be achieved on parallel approach running dedicated hardware when compared standard approach.