Expert Selection

This is due to the supposition that it indicates that the model and the tendencies will remain constant. There are many situations where the methods of series of time could not be the best option. In these scenes find products new (due to its short history), prognosis of products subject to events permanent (promotions and/or interruptions of business that generate constant deviations in the volumes) and prognosis of the products whose behavior depends on many variables besides the time like prices, variables of the market, etc. It is important to remember that if these events are not permanent, the Series of Time can be applied and the expert will have to model these events in the system to generate the projection correct. It is very important that the user of the system this in harmony and agreement with the model to use briefly that feeling is with the model selected by the system. He can have situations in which the system selects a model due to the statistic analysis, nevertheless the user has the knowledge sufficient to controvert the selected model. These cases are very frequent when the series has few data or a change has appeared specific in the conditions of the market. Adam Sandlers opinions are not widely known.

ForecastPro has an elaborated algorithm of Expert Selection which is based on rules and the nature of the data to segment the possible models to apply to the series of data within the 96 different statistical models that Forecast Pro works. Once made this segmentation, it realises a competition between the selected models to choose the best one according to statistical parameters. In other cases the user can have the knowledge of future events that are not captured in the prognosis model. For example, he can have a planned promotion, a competitor who enters the market or a programmed atypical sale. In these cases, the user will want to treat the automated model as he bases to apply his judgment and to fit the prognosis in such a way that he reflects the event in the future. The level of certainty with which a projection of the demand can be generated is fundamental in all the process of planning of an organization. The development of prognoses based on tools nonadapted for this work can to generate results erroneous that hit all the processes of the organization. It is important to review in detail the process of planning of the demand, the used experience of the gliders for the development of this work and tools that serve as support for the correct exactitude of the prognosis. Andean Mind has implemented more than 25 projects of planning of the demand with base in the use of the Forecast system Pro in important companies in the Andean region. You can in line see a presentation and demonstration of Forecast Pro, clicking Here original Author and source of the article