caniasERP Demand Forecasting (DMF) module offers different statistical prediction models in order to plan future needs change, making it an important point in the logistics chan within organizations.
With this module, a demand forecast can be created based on a company's sales figures. This forecast helps decision makers better anticipate future developments and needs and follow a roadmap accordingly. In addition to simple methods such as arithmetic averages, advanced statistical methods such as linear regression analysis and seasonal indexing can also be used in the system to determine the demand forecasting quantities. In addition, with the help of this module, the algorithms that detect and correct the errors in the data set that are the source of demand estimation allow users to predict the future in the most realistic way.
Every organization would want to predict the future in order to take the necessary measures against changing market conditions in a timely manner. Based on this need, it is possible to make plans based on future predictions with the Demand Forecasting module. It is very easy to determine the most suitable forecasting model by evaluating the historical data, make predictions of future sales data using various forecasting models, make rough capacity planning based on these predictions and take the necessary action for the organization on time.
The module uses different prediction models to estimate with the most realistic conditions and offers different estimation options for materials. In this way, it is possible to observe to what extent each possible scenario will affect the forecast with simulation predictions that can be carried out in parallel with the actual process of a material.
There are five different demand forecasting calculation methods in the demand forecasting module. These methods are: reference to old values, mean (arithmetic, harmonic, geometric), exponential correction, linear regression, seasonal indexing.
The methods of data imputation or outlier correction can be used in the Demand Forecastin module. If the Imputation Method is chosen, the system provides a more accurate demand estimation result by filling the periods in which the source data is missing with the selected method when calculating the demand forecasting. These methods are; accept as zero, accept previous, the closest values are average, overall average, median. If the Outlier Correction method is selected, the system checks the accuracy of the source data based on the incorrect data detection method selected while making demand forecasting. At the same time, if it encounters erroneous data, it corrects the erroneous data and provides a more accurate demand forecasting result. The methods of detecting erroneous data in the system are variance test and quarter value width test.
The demand forecasting module works in integration with all modules related to materials, especially Base Data Management, Sales Management, Inventory Management, Production Management, Purchase Management modules. The modules included in the integration provide instant data on all expected inputs and outputs for estimation. Thus, the system is always up to date. Forecasts generated by the demand forecasting module allow decision makers to make more accurate future decisions.