Supply chain | Supply Chain Management, Commerce

Demand Forecasting Algorithm Calculations Equations Are missing

(5) Share

Report

Posted on by Islam Mohab

Dears,

Hope all is well,

i am facing an issue to find the Algorithm Equations for ARIMA , ETS , STL and its combinations.

if anyone can help for that it will be so good as the customer ask frequently about the values shown in the Adjusted demand forecast from where it comes.

thanks,

Islam

I have the same question (0)

All responses (3)

Answers (0)

Sort by

Farouk Galal 55 on at

Like
a
(0)

Report
Copy link

Link copied!

hi eslam those are not equations those are parameters used during forecast generation

Was this reply helpful? Yes No
Suggested answer

Drew Lencsak 435 on at

Like
a
(0)

Report
Copy link

Link copied!

The values that exist are baseline starting values provided by Microsoft. They are intended to be changed to match the configuration your forecast process requires. For example. Seasonality hint of "1" would look for seasonal hints for 1 years period, but if you configure that field to be "12, 3, 1" you'd be looking for yearly, quarterly, and monthly trends. Forecasting model, you would only leave as ARIMA, if that was the statistical model you would like to use for forecast generation. Keep in mind, these are baseline values but can be overridden in the Demand forecasting parameters form as well as specified individually for each Item Allocation Key. This link provides a good overview of all the parameters and configurations and what they are responsible for: Demand forecasting setup - Supply Chain Management | Dynamics 365 | Microsoft Learn

Was this reply helpful? Yes No
Suggested answer

Dogan Adiyaman 698 User Group Leader on at

Like
a
(0)

Report
Copy link

Link copied!
Hi,

I don't think your customer wants to look at the forecasting algorithm and understand the logic.

For example, the general equation for an ARIMA model is:
$Y_{t} = c + ϕ_{1} Y_{t - 1} + \dots + ϕ_{p} Y_{t - p} + θ_{1} e_{t - 1} + \dots + θ_{q} e_{t - q} + e_{t}$
Where:

$Y_{t}$ is the value at time $t$

$ϕ$ are the AR parameters

$θ$ are the MA parameters

$e_{t}$ is the error term

In Dynamics 365, the ARIMA, ETS, and STL models are used by Azure Machine Learning to generate demand forecasts. The Adjusted Demand Forecast values are based on the application of these algorithms to your historical data.

For customization, you can write custom R scripts in Azure ML Studio or Azure ML Designer to modify the behavior of these algorithms.

Your client can compare the results of different models and select the one that best fits their forecasting needs.

Additionally, under Master Planning > Setup > Demand Forecasting Parameters, you can configure how these algorithms behave.

Was this reply helpful? Yes No

Join the conversation

Community site session details

Demand Forecasting Algorithm Calculations Equations Are missing

Helpful resources

News and Announcements

Quick Links

Season of Sharing Community Challenge Winners!

Women in Power Builds Momentum

Congratulations to the June Top 10 Community Leaders

Leaderboard > Supply chain | Supply Chain Management, Commerce

Featured topics

Product updates

Community site session details

Demand Forecasting Algorithm Calculations Equations Are missing

Helpful resources

News and Announcements

Quick Links

Season of Sharing Community Challenge Winners!

Women in Power Builds Momentum

Congratulations to the June Top 10 Community Leaders

Subscribe to this forum!

Select categories

Leaderboard > Supply chain | Supply Chain Management, Commerce

Featured topics

Product updates