Machine Learning as a Service- Challenges and Opportunities

Explore the challenges and opportunities of deploying machine learning models as long-term services in this 24-minute conference talk by Dr. Shou-De Lin, Chief Machine Learning Scientist at Appier. Delve into crucial aspects beyond accuracy and efficiency, including data health, model robustness, objective-goal alignment, and bias management throughout the ML lifecycle. Gain insights into the evolving landscape of Machine Learning as a Service (MLaaS) and learn strategies for ensuring sustainable success in ML model deployment and maintenance.

Syllabus

Introduction
Machine Learning Life Cycle
Machine Learning Ecosystem
Challenges
Data Acquisition
Service Bias
Manual Laboring Bias
Randomization Bias
Evaluation
Service Never Ends
Solution
Conclusion