- Calculating pH of remaining solution after acid-base reaction
- Why the boiling and melting points of n-alkanes are higher than those of iso- or neo-?
- How to evaluate enthalpy variation of combustion of sucrose in a constant volume calorimeter?
- Can corrosion from hydrogen peroxide solution model the corrosion protection activities of Al coating?
- What are the parameters of protonation of a molecule?
- How to get Sodium(Na, Pure) from Cooking Salt (NaCl)?
- How to get back to “enter SIM pin” screen?
- How do I answer questions about my unemployment period?
- What is a polite (not 100%) way to say 'you mind your own business'?
- I have five plus years experience as graphic designer but i done my graduation very recently.
- My contract is ending. Is it acceptable to just leave without any talk with the employer?
- How to deal with spoilers at work?
- Asking for a (level) demotion?
- basic mapping reductions without using Turing machines
- How to systematically think about building implicit tree or a search space graph?
- Count arrays with size n, sum k and largest element m
- Polynomial time approximation of NPC problem
- What is the Name of This Problem?
- Is there an algorithm that can solve chess within the span of a human lifetime?
- Bringing an average student in the lab
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