Long Live the Treasure Hunter
暂无分享,去创建一个
Because they violate common intuitions, there are many ways to interpret the results of novelty search described in the previous chapter. This chapter continues to build the argument that there is no reliable way to consistently achieve particular ambitious goals. The results from novelty search don’t point to a method for solving any particular problem we have, but instead to the frailty of pursuing objectives. Novelty search, like any other search algorithm, cannot magically uncover the solution to any particular ambitious problem. And while this might appear pessimistic, all is not lost: We can still continually discover new and interesting things by collecting stepping stones that may lead to important discoveries in the future, even if they are not predetermined. What this change in philosophy asks of us is to become treasure hunters, accumulating stepping stones without knowing exactly to where they lead, but aware that wherever they lead will likely be interesting and rewarding. Such is the way of the treasure hunter.
[1] Jean-Baptiste Mouret. Novelty-Based Multiobjectivization , 2011 .
[2] David H. Wolpert,et al. No free lunch theorems for optimization , 1997, IEEE Trans. Evol. Comput..
[3] Kenneth O. Stanley,et al. Revising the evolutionary computation abstraction: minimal criteria novelty search , 2010, GECCO '10.