This post is the continuation of Project approach for a Machine Learning project, and it focus on the schedule of a given project on the context of machine learning.
The creation of project schedules is something that can be done if you are able to estimate the efforts and you know the availability of the resources. Without that, all you can build is a ball park estimate that will expose your ingenuity.
If you pretend to perform an estimation of effort in detail by components of the lifecycle without previous experience on the context where you are going to execute, then I’m 99% sure you will fail. This also applies to the project schedule.
Come on!! show me at least a basic project schedule!!!
Ok, here you have something basic, based on the WBS mentioned before:
You have to take into account many hidden activities that are required to be done:
- Security adaptation (encryption, anonymization…),
- Data labelling (identification, and automate data labelling),
- Other automations,
- Size of datasets (direct impact on training executions),