tensorflow / tensorflow/probability
tfp.sts.forcast only applies to future data. Is it possible to "predict" training data?
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
I am using the Structural Time Series of TensorFlow Probability.
I separated my data(1 year) into training data (10 months) and test data (2 months).
After building the model and forecasting, I wondered whether it is possible to "predict" the training data so that I can compare the performance of the model in the training data interval.
Since the arguments of "tfp.sts.forcast" can only specify "num_steps_forecast" which does not include that option, I would ask here.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with tfp.sts.forcast and the Structural Time Series forecasting material; determine whether its inputs support predictions over observed training timesteps rather than only future steps. Define the expected in-sample behavior and comparison output before deciding whether an API change or documentation is needed.
Written by the indexing model from the issue text.
Assessment
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100