23.02.2015 Views

Machine Learning - DISCo

Machine Learning - DISCo

Machine Learning - DISCo

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

R~~~~ss(Frontier, Rule, Literal, &i)<br />

Frontier: Set of literals to be regressed through Rule<br />

Rule: A Horn clause<br />

Literal: A literal in Frontier that is inferred by Rule in the explanation<br />

Oki: The substitution that unijies the head of Rule to the corresponding literal in the explanation<br />

Returns the set of literals forming the weakest preimage of Frontier with respect to Rule<br />

head t head of Rule<br />

body t body of Rule<br />

Bkl t the most general unifier of head with Literal such that there exists a substitution Bli<br />

for which<br />

Ori (Bkl (head)) = Bhi (head)<br />

Return Okl (Frontier - head + body)<br />

Example (the bottommost regression step in Figure 11.3):<br />

h?~~~ss(Frontier, Rule, Literd, @hi) where<br />

Frontier = {Volume(x, us), Density(x, dx), Equal(wx, times(vx,dx)), LessThan(wx, wy),<br />

Weight(y, wy))<br />

Rule = Weight(z, 5) c Type(z, Endtable)<br />

Literal = Weight(y, wy)<br />

6ki = {z/Obj21<br />

head c Weight (z, 5)<br />

body c Type(z, Endtable)<br />

Bhl e {z/y, wy/5], where Bri = (ylObj2)<br />

Return {Volume(x, us), Density(x, dx), Equal (wx, times(vx, dx)), LessThan(wx, 5).<br />

Type(y, Endtable)]<br />

TABLE 11.3<br />

Algorithm for regressing a set of literals through a single Horn clause. The set of literals given<br />

by Frontier is regressed through Rule. Literal is the member of Frontier inferred by Rule in<br />

the explanation. The substitution Bki gives the binding of variables from the head of Rule to the<br />

corresponding literal in the explanation. The algorithm first computes a substitution Bhl that unifies<br />

the Rule head to Literal, in a way that is consistent with the substitution Bki. It then applies this<br />

substitution Oh[ to construct the preimage of Frontier with respect to Rule. The symbols "+" and<br />

"-" in the algorithm denote set union and set difference. The notation {zly] denotes the substitution<br />

of y in place of z. An example trace is given.<br />

explanation applies to only a special case of the target concept. As noted earlier,<br />

for the current example the final rule is<br />

SafeToStack(x, y) t Volume(x, vx) A Density(x, d x ) ~<br />

Equal(wx, times(vx, dx)) A LessThan(wx, 5 ) ~<br />

Type(y, Endtable)<br />

11.2.1.3 REFINE THE CURRENT HYPOTHESIS<br />

The current hypothesis at each stage consists of the set of Horn clauses learned<br />

thus far. At each stage, the sequential covering algorithm picks a new positive

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!