The use of machine learning in the physical sciences has grown immensely over the last several years. While machine learning has been around for decades, recent advances in deep neural networks have contributed to the rapid growth of applications across many disciplines, including the physical sciences. However, many of the traditional benchmark datasets for the development of neural networks are not from the physical sciences, making the use of off-the-shelf machine learning methods not always feasible or practical for the physical sciences. In this course, we will learn the fundamentals of neural networks, from the basics of a neuron to more advanced architectures, and how these tools can be applied to the physical sciences. Important considerations for data preprocessing specific to the physical sciences will be discussed, along with how to evaluate the skill, uncertainty, and confidence of neural networks with specific relevance to the physical sciences.