Hardness amplification in nondeterministic logspace

A hard problem is one which cannot be easily computed by efficient algorithms. Hardness amplification is a procedure which takes as input a problem of mild hardness and returns a problem of higher hardness. This is closely related to the task of decoding certain error-correcting codes. We show amplification from mild average case hardness to higher average case hardness for nondeterministic logspace and worst-to-average amplification for nondeterministic linspace. Finally we explore possible ways of improving the parameters of our hardness amplification results.