Designing with advanced artificial intelligence

The course DBM180 explores the cutting-edge capabilities of machine learning and artificial intelligence for designing innovative systems and experiences. Students will gain a deep understanding of classical ML algorithms and techniques, including supervised, unsupervised learning, and more. They will learn how to apply these powerful methods to tackle complex real-world problems and create AI-powered designs that provide unique value propositions. Specifically, students will employ these methods to design prototypes for solving real-world problems, such as a system to classify products, and analyze movie reviews. The projects will showcase how to harness AI to gain value from diverse data sources. With these hands-on projects and exposure to powerful ML software frameworks (such as Weka), students will gain the knowledge and skills required to push the forefront of AI-powered design. It includes the following:

• A generic architecture for designing intelligent systems comprising data repository, predictive engine, and managing potential feedback loops
• Apply architectural blueprint to design systems (digital or physical) providing predictive response
• Predictive engine uses data to predict optimal system responses to provide optimized experience, help automating workflow, or useful information to the human user
• Collect and store user feedback to further fine-tune and improve the model
• Successful prototypes could lead to further research or publications

• Describe key machine learning algorithms and their implementation

• Conceptualize a phenomenon as a learning problem and identify relevant data

• Recognize core components of ML-enabled systems including outcomes, data, and user experience

• Describe data collection strategies or recognize potential data sources for ML system including training, validation and test data acquisition

• Implement conceived machine learning-powered systems for a given context and problem.

• Distinguish key evaluation metrics to benchmark model performance in a given context

• Develop concepts focused on user experience and prototype its key components in groups or individually culminating in a capstone project

• Reflect individually on learning goals and group contributions

You must meet the following requirements

• The Canvas website users a diversity of supporting materials (such as Weka book on machine learning, example Python notebooks, information on human-AI interfaces, etc.)