I am looking for an overview of stats/ML methods to approach a specific kind of time series problem.
I am looking to predict the next value in a series where (static) explanatory variables are present that explain some of the variance. Ideally, the model will be flexible enough to incorporate possible impact of the static variables on any autoregressive or moving average effects.
The dataset I am training on is small (low 1000s of rolling-window sequences).
The only way I can think of approaching this problem is using recurrent neural networks. I want alternatives that are NOT neural network based.
Desired deliverable is a short write-up in coherent English on a handful of methods that you think are appropriate. Include pros and cons, links to successful implementations on similar problems and the like. Please only bid on this if you are very well-versed in timeseries analysis.
24 freelancers are bidding on average $157 for this job
I suppose ARIMA is suitable for your context. If so, I would like to assist you. I am able to do this analysis in R. Kindly advice me further instructions. Thank you very much. Will be waiting your message...
Hi, I have worked with ARIMA, SARIMA & SARIMAX for timeseries analysis which i believe are suitable for your case. Let me know if you want to discuss further. Regards, Monir