Barclays Interview Questions | Model Selection
Question
You have two models of comparable accuracy and computational performance. Which one should you choose for production and why?
in progress
0
Machine Learning
4 years
3 Answers
1117 views
Great Grand Master 0
Answers ( 3 )
For production, we should choose the one which has less computational performance. As the accuracy are similar, the one with less computation time will have low latency and hence can be used for production.
Assuming higher computational power means less time complexity. For production, it is better to choose higher computational performance. In the world of big data, we need to have algorithms that compute huge volumes with as much less time as possible.
Depends upon many scenarios.
Are they both black box models or one model has a better interpretation than other.
Interpretability can be preferred over black box models.
Training and prediction time can also be considered. one which takes less time to train or to output results
can also be considered.