Approximation Limitations of Pure Dynamic Programming
نویسندگان
چکیده
منابع مشابه
Greedy can also beat pure dynamic programming
Many dynamic programming algorithms are “pure” in that they only use min or max and addition operations in their recursion equations. The well known greedy algorithm of Kruskal solves the minimum weight spanning tree problem on n-vertex graphs using only O(n logn) operations. We prove that any pure DP algorithm for this problem must perform 2 operations. Since the greedy algorithm can also badl...
متن کاملDynamic Policy Programming with Function Approximation
In this paper, we consider the problem of planning in the infinite-horizon discountedreward Markov decision problems. We propose a novel iterative method, called dynamic policy programming (DPP), which updates the parametrized policy by a Bellmanlike iteration. For discrete state-action case, we establish L∞-norm loss bounds for the performance of the policy induced by DPP and prove that it asy...
متن کاملStable Function Approximation in Dynamic Programming
The success of reinforcement learning in practical problems depends on the ability to combine function approximation with temporal di erence methods such as value iteration. Experiments in this area have produced mixed results; there have been both notable successes and notable disappointments. Theory has been scarce, mostly due to the difculty of reasoning about function approximators that gen...
متن کاملContinuous State Dynamic Programming via Nonexpansive Approximation
This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm. Journal of Economic Literature Classifications: C61, C63
متن کاملExtending Dynamic-Programming-Based Plan Generators: Beyond Pure Enumeration
The query optimizer plays an important role in a database management system supporting a declarative query language, such as SQL. One of its central components is the plan generator, which is responsible for determining the optimal join order of a query. Plan generators based on dynamic programming have been known for several decades. However, some significant progress in this field has only be...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: SIAM Journal on Computing
سال: 2020
ISSN: 0097-5397,1095-7111
DOI: 10.1137/18m1196339