I'm trying to build a Time Series RNN with LSTM, with multiple continuous inputs features and the one continuous output, there is opensource code available however I just need to change the code so the algorithm can read from my dataset IE a CSV file as input and then the output being a CSV file it writes to. Most available code has inputs and outputs however does not have a section to provide new inputs it asks to predict. Tensorflow code would be ideal, I really just want the code in a position where I can easily update the hyper-parameters, inputs, output and new inputs to estimated once trained. If Tensorflow is not possible, I'd consider other code.
Your project is a perfect match for my skills. Are you in a major rush with this project or is delivery within one week fine for you. best regards, Joerg.