Applied Neural Networks

3 Credits · Level: Graduate

This course provides a gentle and applied introduction to the theory and practice of artificial Neural Networks (NNs for short). NNs refer to a specific subset of AI and ML algorithms designed to mimic the learning processes of a biological brain. During this course we will explore current NN models and architectures with real world applications. The course begins with a discussion of the NN fundamentals, including the basic mathematical concepts for understanding nonlinearity and feedback loops. We will then transition to topics such as feed forward NNs and multilayered perceptrons, backpropagation and training, regularization and overfitting, SVMs, stochastic networks, and (un)supervised learning. The course continues with a cursory overview of more advanced NN architectures – such as CNNs, RNNs, GANs, and encoders – before ending with a discussion of the substantive, ethical, and policy implications of implementing NNs to solve real life problems. Pre-req: students must have completed an introduction to data science course (ex. PPOL 6803 Intro to Data Science).