- Has any relevant American politician ever admited the Maine was a false flag operation?
- What was the security screening process in Swiss airports, circa 1970?
- Is this picture geniuine, and if so, what is happening here?
- Why do commentators calculate the time periods in Revelation based on a 30 day month?
- What type of data augmentation techniques are used for NLP?
- Can I get the same stats as are on blockchain.info and btc.com by running a full node?
- Theories on renaming: Abraham, Isaac, and Jacob
- What are Software Engineers called in France?
- Applying Photoshop's “Shadow/Highlight” correction using standard image-processing algorithms
- Measuring acoustic phase difference between two receivers
- Could RSA keypairs be used for cryptocurrency addresses?
- Can Secure Multi-party Computation be used to reconstruct secret shares?
- Can I leave the Schengen area and go back if I receive an employee card?
- Help needed to repot my 42" avocado tree in my apt please!
- What is the origin of “tja”?
- German word for feeling suffering from the fact that poor tools negatively impact quality of your work?
- Lilypond - Repeat with Partial Mesaure at the End
- Passage has successive F minor and E minor chords. Is this atonal or chromatic?
- Unique Item Permissions CSOM PowerShell
- Set DocLib Title column in C#
Multi-Source Time Series Data Prediction
I was wondering if anyone has experience with temeseries prediction for data from. multiple sources. So for instance, timeseries a,b,..,z each have their own shape, some may be correlated with others. the ultimate goal is to have a model trained such that the t+1 value for any given data source can be predictied.
I personally have two solutions that in theory could work, but was wondering if anyone knew of other frequently used methods.
1) Multi-task learning with LSTM 2) use feature engineering to model properties of each timeseries source as features along with usual features and use these with LSTM