Applied Machine Learning Systems

This unit explores foundational machine learning concepts, the specialized hardware that accelerates them (GPUs and DLPs), and practical AI application domains including vision, language, robotics, drones, industrial systems, and space operations. Students will build on introductory ideas to preprocess data, train simple models, design prototype integrations, and consider historical and societal impacts to prepare for deeper implementation and capstone work in subsequent units.

Learning Objectives

• Analyze the roles, trade-offs, and performance characteristics of CPUs, GPUs, and Deep Learning Processors (DLPs) for training and deploying machine learning models, using measurable metrics such as throughput, latency, and energy consumption for given use cases
• Apply core machine learning principles—probability, neural network basics, supervised/unsupervised learning, and data preprocessing—to build, train, and evaluate a simple model using GPU/DLP-accelerated tooling and report model performance and data-cleaning procedures
• Demonstrate integration of AI capabilities into applied systems by designing and prototyping workflows for computer vision, natural language interfaces, and basic robotics/drones, specifying sensors, models, data flows, and human-AI interaction points
• Evaluate domain-specific and societal considerations for AI deployment (industrial automation, mobility, and space), including historical lessons such as AI winters, and propose strategies to address safety, reliability, workforce impacts, and ethical concerns