Imaginghub Forum - ML price prediction problem on time series data

I have some time series data describing purchasing logs of various items, like product P purchased for amount M on D date at Z place, and some other additional features. All I want to do is to predict how much does it need to be paid for purchasing P product on the date D+n at Zn place with identical additional features. How to approach this? I'm thinking of using Tensorflow.

Answers

Just throwing data at TF and expecting to get some result isn't a good idea.

Which python library is better to use in your opinion then?

The key is not about the specific library. First, you should do some analytical work on your data, make sure that your features are meaningful and have the same and right format, look at the distributions and descriptive statistics and then decide whether MLR is a good approach.

Add Answer

ML price prediction problem on time series data

ML price prediction problem on time series data

Answers

Need support?