Tudnivalók az Y0-modell offline kialakítás kapcsán

Steps/recommendations:

  • The aim of Y0 models is to make analysis in case of simulated (each object is identical/constant) Y-vectors, which, if run correctly, will show if there is such a view of the input-variables, that makes the differences of the Y disappear.
  • The Y0 model is the approximation (cf. Dobó: Joker) of the universal (lacking tendencies) interpretation of similarity, its intent is to find out if there is stair function that assures this kind of universality (so, the expert systems).
  • In case of public procurement task that are already in progress, the Y0 provides that price is among the X-variables, and besides that, the importance proportions of the X-variables (variable groups) can be given subjectively.
  • In this case, the results have to be interpreted (e.g. in case where Y=1) as searches for objects, where the estimation value is either higher or smaller than 1. In the former case, an offer is a relative winner, while in the latter case it is a loser of the offering-evaluation. The higher the distance of an object's value from 1 (where this object is higher than 1), the more advantageous is the offer.
  • When determining the winner, it should be taken into consideration, that the proportion of estimation and the real price makes the winner relative in economic sense, because the prices assigned to relative advantages makes a difference. In the Y0 model, the estimation without dimension does not allow this. Even in case of the standard models, we can only talk about the 'it is not better, yet it is still more expensive'-type price advantage. Real price advantage is not necessarily an element of the COCO STEP IX solution, because in many cases, an object may worth any price, taking the other offers into consideration.
  • When using the similarity analysis for policy-orientation (strategic level support, e.g. regional development), the facts of context-free losses of balance can be pointed out with the help of an Y0 model: e.g. is there a disadvantageous situation in case of certain objects, based on a given data asset...
  • For Y0 models, it is advised to specify a distance more than zero between two stairs in the auxiliary matrix: (demo).

Should you have any further questions, we suggest you to take a look on the related documents of MIAU: e.g. Demo


((Back))