
Lab: Image Clustering & Classification
Session with George Anadiotis (Founder, Principal Consultant, Linked Data Orchestration)
About this Session
Unlock the Business Value of Visual Data
Step into the world of computer vision in this hands-on workshop where we'll transform unstructured visual data into business outcomes. Working with the image dataset introduced earlier, we'll discover how organizations extract actionable intelligence from photos, scans, and other visual assets without requiring specialized expertise.
Experience firsthand how algorithms automatically organize images into meaningful groups based on visual similarities, revealing patterns human analysis may miss. Learn how feature engineering works differently with images as we leverage both accompanying metadata and powerful embeddings, the numerical representations that capture visual meaning.
Develop practical classification models that can automatically categorize incoming images, potentially automating workflows that previously required manual review. This session bridges the gap between theoretical understanding and practical application of visual AI, providing essential experience for professionals considering image-based machine learning solutions in their organizations.
- Outline
- Exploring options
- Choosing an algorithm
- Feature engineering
- Using metadata
- Using embeddings
- Development, validation and testing
- Evaluation
- Format
- This session is hands-on. Attendees work on their laptops
- Level
- Intermediate
- Prerequisite Knowledge
- Machine Learning Algorithms & Evaluation
- Learning Outcomes
- Experience developing machine learning models using metatada, embeddings, and multi-modal data