Title : 
Bounded Independence Fools Halfspaces
         
        
            Author : 
Diakonikolas, Ilias ; Gopalan, Parikshit ; Jaiswal, Ragesh ; Servedio, Rocco A. ; Viola, Emanuele
         
        
            Author_Institution : 
Dept. of Comput. Sci., Columbia Univ., New York, NY, USA
         
        
        
        
        
        
            Abstract : 
We show that any distribution on {-1,+1}n that is k-wise independent fools any halfspace (a.k.a. threshold) h : {-1,+1}n ¿ {-1,+1}, i.e., any function of the form h(x) = sign(¿i=1 n wiXi - ¿) where the w1,..., wn, ¿ are arbitrary real numbers, with error ¿ for k = O(¿-2 log2(1/¿)). Our result is tight up to log(1/¿) factors. Using standard constructions of k-wise independent distributions, we obtain the first explicit pseudorandom generators G : {-1,+1}s ¿ {-1,+1}n that fool halfspaces. Specifically, we fool halfspaces with error e and seed length s = k · log n = O(log n · ¿-2 log2(1/¿)). Our approach combines classical tools from real approximation theory with structural results on halfspaces by Servedio (Comput. Complexity 2007).
         
        
            Keywords : 
Boolean functions; approximation theory; computational complexity; log normal distribution; random number generation; Bounded Independence Fools Halfspaces; approximation theory; distribution function; explicit pseudorandom generators; k-wise independent fools; Approximation methods; Boosting; Circuits; Computer science; Educational institutions; Game theory; Information science; Silicon; Support vector machines; Voting; halfspaces; k-wise independent distributions; pseudorandomness;
         
        
        
        
            Conference_Titel : 
Foundations of Computer Science, 2009. FOCS '09. 50th Annual IEEE Symposium on
         
        
            Conference_Location : 
Atlanta, GA
         
        
        
            Print_ISBN : 
978-1-4244-5116-6
         
        
        
            DOI : 
10.1109/FOCS.2009.68