Machine Learning Canvas

Machine Learning Canvas

Project phase: Business and Data Understanding

You translate the business problem into a concrete machine learning problem on a single structured canvas: the prediction task and its output, the decision it supports and who makes it, the value for the end user, the data and features available, how often predictions are made, and how you'll evaluate the system both offline and in real use. Every block is filled with project-specific, mutually consistent content, and you state the assumptions that would break the design if they turned out false.

Starting Points

• Machine Learning Canvas

Key Points

• The canvas is filled with project-specific content (no generic placeholders).
• All blocks are logically consistent (e.g., the data you list can actually support the prediction task you defined).
• The core problem is clearly distinguished from secondary issues and the societal context is addressed.
• Assumptions that could break the design are explicitly stated.
• The deliverable is a standalone markdown document, made available in GitLab.