Going back to the topic, this fall is my last semester in my Master of Business Analytics program in USC, and the core tool being used in this program is R. Since I am taking one very interesting time-series analysis course taught by Professor Gabrys, I have thought about whether to write an article to talk about it, as well as wrapping up my study in USC. Now it is the chance for me, because I just summed up the basic ARIMA analysis in R, with the procedure taught again and again by my professor. This is an automated function that can help users to come up with candidate ARIMA models for the dataset.
For people who have no background in time-series data analysis or in ARIMA process, it might be a bit overwhelming for you at the beginning. It requires the basic understanding in time-series data analyzing process. This function, in which I mimicked how my professor emphasizes time and time again in class, will provide users with another way to explore potential good ARIMA models, despite of the fact that R has a lot of automated functions that do it all without giving the users much information.
The code, and the detailed manual are up on my Github. This perhaps is going to be my last share in automation as a USC student, and I have been always proud to be a Trojan.
Automation, on the other hand, is the very initial purpose of my blog, and one thing I have learned in my final days on campus is, automation, AKA, computer programs, shall never replace human minds. instead, we as humans, should learn how to use machines to enhance the capabilities and overcome the shortcomings of our brains — period.