Firstly, we have no knowledge about advanced data analysis or data mining.

We are working with process plant which gather data that comes into the process plant.
We use sensors data for the input to the process plant such as flow rates and concentration of nutrients (minutely time series data).

We plan to develop prediction (hourly) to our time series. We are discussing if using PCA (Principal Component Analysis) can help us to simplify the variables in prediction. However, I am not sure how PCA can give information for prediction. What do you think? Is there any better way to approach this objective?


Transfer Functions can be very useful in analyzing this kind of time series data. Take a look at http://www.autobox.com/cms/index.php/news/48-autobox-software-review-in-orms , particularly the material on the furnace data analysis (p68) . Process control is a very important application of time series causal analytics and this review is very informative in this regard. For a general introduction to transfer function analysis see https://onlinecourses.science.psu.edu/stat510/node/75

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.