Randomized Algorithms (Spring 2010)/Fingerprinting

Fingerprinting

Checking identities

Example: Checking matrix multiplication

Consider the following problem:

• Given as the input three [math]\displaystyle{ n\times n }[/math] matrices [math]\displaystyle{ A,B }[/math] and [math]\displaystyle{ C }[/math],
• check whether [math]\displaystyle{ C=AB }[/math].

Algorithm (Freivalds) pick a vector [math]\displaystyle{ r \in\{0, 1\}^n }[/math] uniformly at random; if [math]\displaystyle{ A(Br) = Cr }[/math] then return "yes" else return "no";

If [math]\displaystyle{ AB=C }[/math] then [math]\displaystyle{ A(Br) = Cr }[/math] for any [math]\displaystyle{ r \in\{0, 1\}^n }[/math], thus the algorithm always returns "yes".

Lemma If [math]\displaystyle{ AB\neq C }[/math] then for a uniformly random [math]\displaystyle{ r \in\{0, 1\}^n }[/math], [math]\displaystyle{ \Pr[A(Br) = Cr]\le \frac{1}{2} }[/math].

Example: Checking polynomial identities

Consider the following problem:

• Given as the input two multivariate polynomials [math]\displaystyle{ P_1(x_1,\ldots,x_n) }[/math] and [math]\displaystyle{ P_2(x_1,\ldots,x_n) }[/math],
• check whether [math]\displaystyle{ P_1\equiv P_2 }[/math].

Algorithm (Schwartz-Zippel) pick [math]\displaystyle{ r_1, \ldots , r_n }[/math] independently and uniformly at random from a set [math]\displaystyle{ S }[/math]; if [math]\displaystyle{ P_1(r_1, \ldots , r_n) = P_2(r_1, \ldots , r_n) }[/math] then return “yes” else return “no”;

Theorem (Schwartz-Zippel) Let [math]\displaystyle{ Q(x_1,\ldots,x_n) }[/math] be a multivariate polynomial of degree [math]\displaystyle{ d }[/math] defined over a field [math]\displaystyle{ \mathbb{F} }[/math]. Fix any finite set [math]\displaystyle{ S\subset\mathbb{F} }[/math], and let [math]\displaystyle{ r_1,\ldots,r_n }[/math] be chosen independently and uniformly at random from [math]\displaystyle{ S }[/math]. Then [math]\displaystyle{ \Pr[Q(r_1,\ldots,r_n)=0\mid Q\not\equiv 0]\le\frac{d}{|S|}. }[/math]

Fingerprinting

Suppose we want to compare two items [math]\displaystyle{ Z_1 }[/math] and [math]\displaystyle{ Z_2 }[/math]. Instead of comparing them directly, we compute random fingerprints [math]\displaystyle{ \mathrm{FING}(Z_1) }[/math] and [math]\displaystyle{ \mathrm{FING}(Z_2) }[/math] and compare these. The fingerprints has the following properties:

• If [math]\displaystyle{ Z_1\neq Z_2 }[/math] then [math]\displaystyle{ \Pr[\mathrm{FING}(Z_1)=\mathrm{FING}(Z_2)] }[/math] is small.
• It is much more to compute and compare the fingerprints than to compare [math]\displaystyle{ Z_1 }[/math] and [math]\displaystyle{ Z_2 }[/math] directly.

For Freivald's algorithm, the item to compare are two [math]\displaystyle{ n\times n }[/math] matrices [math]\displaystyle{ AB }[/math] and [math]\displaystyle{ C }[/math], and given an [math]\displaystyle{ n\times n }[/math] matrix [math]\displaystyle{ M }[/math], the random fingerprint is computed as [math]\displaystyle{ \mathrm{FING}(M)=Mr }[/math] for a uniformly random [math]\displaystyle{ r\in\{0,1\}^n }[/math].

Universal hashing

Example: checking distinctness