I have this problem I don't know how to solve. I have X variables. One of them is a key value. I have different data sets, each containing rows of key values and values corresponding to 1 or more of the other variables. I have a confidence value for each data set, and confidence values for each data row in the set. How do I best choose a value for each of the non-key variables?

e.g.:

Variables: Key - Element_1 - Element_2 - Element_3

Data from Data Set 1 (confidence of data set: 9/10):
K1 - 1 - NULL - 3 (confidence of this row: 7/10)
K2 - 2 - NULL - 4 (confidence of this row: 8/10)

Data from Data Set 2 (confidence of data set: 7/10):
K1 - 5 - 6 - NULL (confidence of this row: 8/10)
K2 - 7 - 8 - NULL (confidence of this row: 9/10)

Now we would choose 1 as the value of Element_1 for key K1 because we have a confidence score of 9 * 7 = 63 from Data Set 1, whereas we only have a 7 * 8 = 56 confidence score for choosing 5 as the value of Element_1 for key K1 using Data Set 2.

Thank you.