I have a working example of seq2seq RNN forecasting code, it is working well on 1 step prediction. However, when I tried to predict more than 1 step it failed. the problem is with the shape and probably need more tweaking on the architecture. all training and validation is already there, probably add R2 and RSME is better. for now i use mae and smae. It's all build in Keras.
7 freelancers are bidding on average €94 for this job
I am a professional data scientist from Scotland, I have a great amount of experience in multivariate analysis and RNN forecasting. I am here to discuss your project further.
Hello. I have read your project details before placing my bid. I can assure you that I can deliver high quality work on time. I am willing to start immediately. Thank you very much.
hello, I'm very interested in your project. I worked on several prediction/machine learnig projects with python keras .Please contact me to discuss more about the project.