Context and scope

TensorFlow, an open-source machine learning framework developed by the Google Brain team, was initially released in 2015, quickly gaining widespread adoption. TensorFlow provides a comprehensive ecosystem for building and training machine learning models across various domains, from computer vision and natural language processing to reinforcement learning and more.

Goals and non-goals

Machine Learning with TensorFlow JS is a beginner-friendly course. This course is designed to give students exposure to recent ai developments and a fundamental understanding of machine learning paradigms. Unlike traditional courses taught, a deep-dive on the underlying mathematics will not take place and the final projects will be a communicable, portable project.

Design

A inverted class format will take place from most to least recent. Simpler models with greater user control will be used as the course progresses.

Data Storage

A single, monolithic git repository will store both instructional material and demonstration projects.

Degree of Constraints

Limited knowledge of beginners on Cloud computing, server-side development makes deploying models trained through Python or C++ versions of TensorFlow difficult.

TensorFlow JS’s weak performance on larger model sizes is a significant limitation attempted problems. Ease of install and learnability are a significant advantage.

Alternatives Considered

PyTorch is a logical choice for AI and Machine Learning given recent growth. However, PyTorch’s limited industry adoption, lack of JavaScript port and fewer deployment options discouraged PyTorch’s use. Concepts taught are transferable to PyTorch.

Cross-cutting concerns

Keeping track of participation and process can not be achieved by a single markdown book. Slack will be used for administration and a notion page will be used to track progress.