This im­portant, yet pre­vious­ly under-​addressed ques­ti­on is the focus of a newly pu­blished ar­ti­cle in the Jour­nal of Ar­ti­fi­cial In­tel­li­gence Re­se­arch. Toge­ther with col­le­agues from Ire­land and the US, CSBA di­rec­tor Dr. Chris­ti­an Haas shows that dif­fe­rent im­pu­ta­ti­on stra­te­gies can have a si­gni­fi­cant ef­fect on the re­sul­ting fair­ness of pre­dic­tions, rai­sing awa­ren­ess that Al­go­rith­mic Fair­ness con­side­ra­ti­ons need to be con­side­red in all steps of a Data Ana­ly­tics and Ma­chi­ne Lear­ning pipe­line.

Read the (open ac­cess) ar­ti­cle here.