A new disaggregation preference method for new products design

Cognitive Effort Reduction Within Group Decision Making Through Aggregation and Disaggregation of Individual Preferences

“…Indeed, depending on the application domain, wrong and missing assignments may have different consequences. In domains where 'no decision' is better than 'wrong decision' as in human health, we should define α such that α < 1 2 , which penalizes wrong assignments more than missing assignments. Defining α such that α > 1 2 , gives more penalty to missing assignments than wrong assignments.…”

“…In domains where 'no decision' is better than 'wrong decision' as in human health, we should define α such that α < 1 2 , which penalizes wrong assignments more than missing assignments. Defining α such that α > 1 2 , gives more penalty to missing assignments than wrong assignments. Setting α = 1 2 gives the same penalty to missing and wrong assignments.…”

“…Defining α such that α > 1 2 , gives more penalty to missing assignments than wrong assignments. Setting α = 1 2 gives the same penalty to missing and wrong assignments.…”

“…Each student has been assigned by the decision makers into a range [l i , u i ] (i = 1, 2, 3) where l i and u i are the lower and higher class where he/she may be assigned according to decision maker d i . Typically, each student is assigned by default to the range [1,3]. If the decision maker changes these values, then the student's assignment will be constrained and it becomes an assignment example.…”

“…Then, an inference procedure is used to deduce, through a certain form of regression, a set of preference parameters values that re-do at best the assignment examples. Aggregation/disaggregation approach has been used largely for single decision makers [1][7] [10][16] [18]. There are also some group oriented aggregation/disaggregation approaches [2][6] [15].…”