|
Title:
|
TOWARDS EXPLAINABLE MACHINE LEARNING OPERATIONS (MLOPS) |
|
Author(s):
|
Krzysztof Dzieciolowski and Ningsheng Zhao |
|
ISBN:
|
978-989-8704-50-4 |
|
Editors:
|
Piet Kommers, Mário Macedo, Guo Chao Peng and Ajith Abraham |
|
Year:
|
2023 |
|
Edition:
|
Single |
|
Keywords:
|
Machine Learning Operations (MLOps), Model Explanations, Model Diagnostics, Importance Sampling |
|
Type:
|
Reflection |
|
First Page:
|
384 |
|
Last Page:
|
387 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
In the dynamically expanding AI applications, there is a need for an effective operational framework for managing ever increasing number of predictive models. A new field of Machine Learning Operations (MLOps) has recently emerged to address the issue of efficient operations and management of machine learning models. However, enterprises still struggle to develop effective MLOps systems due to lack of expertise and limited experience in AI operational needs. Attempts of an ad-hoc administration of ML algorithms have not been successful to deal with the challenges of fast proliferation of AI models. At the same time, existing concepts of MLOps frameworks don't adequately address an important question of accurate machine learning models diagnostics and explainability. In this reflection paper, we review commonly used MLOps approaches and explainability methods and suggest novel methods to address challenges of current model explainability methods. We discuss the need for university curriculum to address the issues facing management of AI models, their diagnostics and explainability. |
|
|
|
|
|
|
