Neural networks on a micro controller - Preparation and math
I've been playing around with the idea of running a neural network on a micro controller. This blog post expects the reader to have knowledge of neural networks. For more information, you can check out the following links: http://neuralnetworksanddeeplearning.com/ http://www.deeplearningbook.org/contents/intro.html https://uk.mathworks.com/help/nnet/ug/improve-neural-network-generalization-and-avoid-overfitting.html?w.mathworks.com This project is not designed to be the most efficient neural network, but rather one that runs on a micro controller with very limited resources. The goal The goal is to make a robot on the end of a stick. The robot will have several motorized joints and potentiometers, and will be attached to a center point with a long stick. The center point also has a rotary encoder which measures the angle. The robot will have a single blue pill controlling it. The micro controller has the following specs: 72 MHz clock speed 64 KB flash 20 KB R