Online Learning:Formulation At each round t=1,2,... (1)the player first picks a model wW; (2)and simultaneously environments pick an online function fr:W->R; (3)the player suffers loss ft(wt),observes some information about ft and updates the model. ·An example of online function f:W→R. Considering the task of online classification,we have (i)the loss e:Jy×Jy→R,and fi(w)=l(h(w;x:),) (i)the hypothesis function h:W×X→). =e(w xt,Ut)for simplicity Advanced Optimization(Fall 2023) Lecture 5.Online Convex Optimization 8Advanced Optimization (Fall 2023) Lecture 5. Online Convex Optimization 8 Online Learning: Formulation for simplicity • Considering the task of online classification, we have