SDS 578: Identifying Commercial ML Problems — with Jon Krohn

Self-Reflection

Jon Krohn emphasizes the importance of identifying a commercial problem before diving into data collection or machine learning model development. He suggests that self-reflection can be a powerful tool in this process, encouraging practitioners to ask themselves if there are processes in their personal or professional lives that could be automated. This reflective approach can lead to innovative ideas for machine learning applications.

Reflect, reflect, reflect. And by writing it out like that every morning, you'll hopefully cue the question more often.

--- Jon Krohn

By consistently pondering this question, one might discover opportunities for automation that were not initially apparent 1.

   

User Feedback

Engaging with power users and prospective clients is another strategy Jon highlights for identifying machine learning problems. He notes that these interactions can reveal missing functionalities or areas for improvement that can be addressed with machine learning. Prospective clients, in particular, often provide valuable insights by comparing your product with competitors, which can inspire new machine learning functionalities.

Prospective clients will often deliberately contrast you with your competitors, which can generate direct or indirect ideas as to novel machine learning based functionality.

--- Jon Krohn

These insights can be crucial for developing competitive and innovative machine learning solutions 2 3.