<div dir="auto">Here's a Banks efficient version that more closely mimics our proven Landau efficient method.<div dir="auto"><br></div><div dir="auto">First, for comparison a slight reformulation of Agenda Based Landau:</div><div dir="auto"><br></div><div dir="auto">Definition: For each agenda item X, let thrall(X) be the set of agenda items defeated by X.</div><div dir="auto"><br></div><div dir="auto">Initialize X as the most favorable item on the agenda. </div><div dir="auto"><br></div><div dir="auto">While some agenda item Y defeats both X and every member of thrall(X), let the most favorable such Y be the new value of X. </div><div dir="auto"><br></div><div dir="auto">Now for the analogous Banks efficient method:</div><div dir="auto"><br></div><div dir="auto">Initialize X as the most favorable agenda item. Then while some agenda item Y defeats both X and every member of chain(X), replace X with the most favorable such Y.</div><div dir="auto"><br></div><div dir="auto">To complete the definition, we need to define chain(X):</div><div dir="auto"><br></div><div dir="auto">Chain(X) is the covering chain of thrall(X) constructed by the following procedure ...</div><div dir="auto"><br></div><div dir="auto">Initialize 'chain' as the empty set. Then ...</div><div dir="auto">While some Z in thrall(X) defeats every member of 'chain', incorporate the most favorable such Z into 'chain'. EndWhile.</div><div dir="auto"><br></div><div dir="auto">As you can see, the difference between the two methods is that the rôle of thrall(X) in the first method is taken over in the second method by a set chain(X) that covers thrall(X).</div><div dir="auto"><br></div><div dir="auto"><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">El mié., 27 de jul. de 2022 2:30 p. m., Forest Simmons <<a href="mailto:forest.simmons21@gmail.com">forest.simmons21@gmail.com</a>> escribió:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto">This procedure is composed of two sub-procedures ... one to construct an "advanced placement" chain headed by the most favorable agenda item ... and another to complete the chain by a process of chain climbing.<div dir="auto"><br></div><div dir="auto">First sub-procedure: initialize a chain 'c' as the empty set. Then while there exists (even vacuously) some agenda item that is defeated by every member of 'c', incorporate the most favorable such item into 'c'.</div><div dir="auto"><br></div><div dir="auto">Final sub-procedure: while any agenda item defeats every member of 'c', incorporate the most favorable such item into 'c'.</div><div dir="auto"><br></div><div dir="auto">After executing both sub-procedures, elect the head of the finished chain.</div><div dir="auto"><br></div><div dir="auto">This method is definitely Banks efficient. I'm pretty sure that it is also monotonic and (unlike ordinary chain climbing) Independent from Pareto Dominated Alternatives (IPDA).</div><div dir="auto"><br></div><div dir="auto">I'll send proof sketches soon (if it continues to hold up).</div><div dir="auto"><br></div><div dir="auto">-Forest</div><div dir="auto"><br></div><div dir="auto"><br></div></div>
</blockquote></div>