Professional Documents
Culture Documents
Quality Assurance and Testing The ML Model - DZone AI
Quality Assurance and Testing The ML Model - DZone AI
Quality Assurance and Testing The ML Model - DZone AI
LIVE WEBINAR - Thursday, May 14th: Build Your Own CDN in Under 1 Hour
Sign Up Now
DZone > AI Zone > Quality Assurance and Testing the ML Model
This is the irst post in the series of posts related to Quality Assurance (QA) and Testing Practices and
Data Science/Machine Learning Models, which I will release in the next few months. The goal of this
and upcoming posts is to create a tool and framework that can help you design your testing/QA
practices around data science/Machine Learning models.
If the answers to the above questions are yes, then keep reading. I will be presenting concepts, tools,
and frameworks that will help you achieve some of the objectives mentioned earlier.
I have seen in my experience that ML models are developed and tested by data scientists themselves.
This is not a desired situation to be in. Ideally speaking, it should be a quality assurance team that
should be performing QA by running tests as like traditional software to test the ML models from time-
to-time. However, the challenge is that ML models are not like traditional software where the behavior
of the software is pre-determined based on the different inputs. We will touch upon some of the
challenges related to testing ML models in later articles.
Quality of data
Quality of features
Quality of ML algorithms
https://dzone.com/articles/quality-assurancetesting-the-machine-learning-mode 1/3
07/05/2020 Quality Assurance and Testing the ML Model - DZone AI
The role of the QA is to put test mechanisms in place to validate whether the data used for training is
sanitized. In other words, the tests need to be performed to identify whether there are instances of data
poisoning attacks intentionally or unintentionally.
In order to achieve the above, one of the techniques could be to have QA/Test engineers work with
product management and product consultant teams for some of the following:
Understand the statistics related with data (mean, median, mode etc)
Understand the data and their relationships at a high-level
Build tests (using scripts) to check the above statistics and relationships.
Run the tests at regular intervals
The parameters listed above would need to be tracked at regular intervals and veri ied with the help of
PMs/consultants before every release. We will go into the details in later articles.
One of the ways to go about testing ML algorithms with new data is the following:
Keep all the ML models based on different algorithms handy. Many times, I have seen that ML
d l b l d ff l h d
https://dzone.com/articles/quality-assurancetesting-the-machine-learning-mode d d d df ll f h 2/3
07/05/2020 Quality Assurance and Testing the ML Model - DZone AI
models are built using different algorithms and get discarded once and for all after the most
accurate model gets selected.
Retrain all of the models and evaluate the performance
Track the performance of all the models with new data set at regular intervals.
Raise the defect if another model starts giving greater accuracy or performing better than the
existing model.
References
Keeping Your Machine Learning Models Up-To-Date
An introduction to feature selection
Summary
In this post, you learned about the need for QA practices for Data Science/ML models and also the
different aspects of testing the ML models. Please feel free to suggest or share your thoughts in the
comments section.
Topics: ARTIFICIAL INTELLIGENCE , DEEP LEARNING , MACHINE LEARNING , ML ALGORITHMS , ML MODEL , QUALITY ASSURANCE ,
TESTING ML MODELS
Published at DZone with permission of Ajitesh Kumar , DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.
https://dzone.com/articles/quality-assurancetesting-the-machine-learning-mode 3/3