Questions tagged [mlops]
The mlops tag has no usage guidance.
7
questions
1
vote
1
answer
45
views
How to automatically trigger model retraining for object detection models in case of data drift?
I have fine-tuned a pre-trained Yolov8 model on my dataset of labelled containers in a warehouse conveyor belt (image example here) .
I am working to develop an MLOps project for my project portfolio, ...
- 40.5k
0
votes
2
answers
27
views
How to filter out class for which the model has not been trained in ml web app?
I have developed an python based ml web app. It gives details of the book from a image of book cover.
Problem: When I upload the book cover image then it works but when i click image of any random ...
- 206
0
votes
0
answers
18
views
model versioning and validation of online models
While training an online model, usually we use progressive validation https://maxhalford.github.io/blog/online-learning-evaluation/. My question is what we should do when we detected performance ...
- 101
0
votes
1
answer
65
views
How can batch prediction make drift monitoring easier than online prediction?
In this video, I learned that drift monitoring would be easier in batch prediction than that in online prediction:
But I don't know why and I cannot find any information about it googling. In my ...
CommunityBot
- 1
3
votes
2
answers
80
views
Why business experts should prefer state-of-the-art deep neural networks over simpler models? [closed]
I have encountered this pattern for a long time (5+ years). So many professionals come with an interesting domain-specific problem, and they demand using state-of-the-art deep learning models: take it ...
- 1,174
-1
votes
1
answer
121
views
When is it better to utilize machine learning over heuristics?
I learned that 87% of machine learning projects fail due to these five pitfalls:
the scope of the project is too big;
the project’s scope increased in size as the project progressed—e.g., scope creep;...
- 907
2
votes
2
answers
184
views
How will MLOps and lifelong learning be complementary?
According to [1], in MLOps, continuous training is
a new property, unique to ML systems, that's concerned with automatically retraining and serving the models.
While lifelong/incremental learning ...
- 907