Undefinable True Target Learning
Yongquan Yang · Qeios · 2024
The situation where the learning true target cannot be precisely defined is quite common in various artificial intelligence (AI) application scenarios. Yet, this situation has not been systematically analysed. In this article, we formally refer to this situation as undefinable true target learning (UTTL). From the perspectives of problem definition, alternative solution, specific method, and particular application, we present the first fundamental basis for systematically analysing the UTTL situation in AI application scenarios.