We're back! For our first meetup of 2018 we're happy to present Andrew Morgan, author of Mastering Data Science with Apache Spark. Andrew is an experienced data scientist and platform engineer and will talk about “Deep Training” of Genetic Algorithms for IoT; using Tensorflow accelerated Evolutionary Algorithms in python, for deploying ML to IoT edge devices.
Andrew will present the use of Karoo_GP, a python program that leverages TensorFlow to accelerate machine learning via genetic algorithms. He will walk through a use case, and highlight benefits of the technique to the IoT community.
Until recently GEP (Gene Expression Programming) a spin-off genetic / evolutionary ML approaches have had good success for machine learning problems where training data sets have been reasonably small in size. Karoo_GP extends the practical use of these techniques to larger datasets; it does this by expressing the evolved functions as tensorflow operations for evaluation on GPUs. The approach has benefits specifically for the IoT community, as the models created are highly portable, and are easily deployable to constrained hardware environments.
Teas, coffees, cold drinks and subway sandwiches provided!