0
$\begingroup$

I'm working on a project where I have to predict the total quantities sold at the ITEM/DAY level.

As a model, I thought of using an ARIMA (I'm using R), but I wanted to get a confirmation of whether it is a good choice or not. If not, can I get the contexts where the ARIMA model is the most useful?

Improve this question
$\endgroup$

2 Answers 2

Reset to default
1
$\begingroup$

A popular and widely used statistical method for time series forecasting is the ARIMA model. ARIMA is an acronym that stands for AutoRegressive Integrated Moving Average. It is a class of model that captures a suite of different standard temporal structures in time series data. You can read about it more here.

Based on your problem statement, it definitely looks like ARIMA model can help as you are predictions are based of previous time series data. However, as a rule of thumb in Machine learning, there is no one model that can solve the problem out of the box. There is a saying in statistics that goes something like this :

All models are wrong, some are useful.

What this implies is, you can start with ARIMA but you need evaluate your model to see if ARIMA can solve the problem for you or you need to explore further.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thank you for the precision. $\endgroup$
    – ilni
    Commented Jun 1, 2019 at 13:00
0
$\begingroup$

Its most useful where your data fits an autoregressive integrated moving average model. You test this using the standard statistical tests for time series analysis. Consult a standard text on time series analysis for more information.

Improve this answer
$\endgroup$
1
  • $\begingroup$ thank's for the insight. $\endgroup$
    – ilni
    Commented Jun 1, 2019 at 13:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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