This document discusses developing and testing a neural network model to perform classification tasks on electronic signal data. It describes preprocessing the raw data using fast Fourier transformations, designing a fully connected neural network with one hidden layer of 15 nodes, training it for 700 epochs with 2000 passes to achieve over 84% accuracy, and testing it on FFT data with limited success due to the small testing data set. The conclusion notes poor overall test results of 33% accuracy, and that further improvements could involve a multi-layered or partially connected network topology with more testing of parameters and data sets.