Time-inconsistent planning

@article{Kleinberg2018TimeinconsistentP,
  title={Time-inconsistent planning},
  author={Jon M. Kleinberg and Sigal Oren},
  journal={Communications of the ACM},
  year={2018},
  volume={61},
  pages={99 - 107}
}
In many settings, people exhibit behavior that is inconsistent across time---we allocate a block of time to get work done and then procrastinate, or put effort into a project and then later fail to complete it. An active line of research in behavioral economics and related fields has developed and analyzed models for this type of time-inconsistent behavior. Here we propose a graph-theoretic model of tasks and goals, in which dependencies among actions are represented by a directed graph, and a… 

Figures from this paper

Planning with Multiple Biases
TLDR
This work proposes a theoretical model for planning with a pair of biases exhibiting an intuitively natural interaction: present bias and sunk-cost bias, and shows how certain natural behavioral phenomena can arise in this model only when agents exhibit both biases.
The Price of Uncertainty in Present-Biased Planning
TLDR
The graphical model of Kleinberg and Oren is used and it is shown that the conceptual loss of efficiency incurred by incentives in the form of penalty fees is at most $2$ in the former and $1 + \max B/\min B $ in the latter model.
Learning Time Dependent Choice
TLDR
It is shown that if the learner is given full power over the data he learns from -- in the form of learning via membership queries -- even very naive algorithms significantly outperform the guarantees provided by higher level active learning algorithms.
Reinforcement learning and inverse reinforcement learning with system 1 and system 2 Anonymized for Review
  • Computer Science
TLDR
A simple IRL algorithm is adapted to correctly infer the goals of dual system decision-makers and allows us to make interventions that help, rather than hinder, the dual-system agent’s ability to reach their true goals.
Complexity and Algorithms of K-implementation
TLDR
This paper settles the complexity of K-implementation, a ten-year open problem in AI, by proving that the problem is NP-COMPLETE for general games with respect to dominance by pure strategies and providing an optimal algorithm based on mixed-integer linear program.
2 9 M ay 2 01 9 Directed Graph Minors and Serial-Parallel Width
TLDR
This work characterizes all TDAGs with serial-parallel width at most k; a class of networks known to guarantee bounded negative externality in nonatomic routing games; and proposes new variants of directed graph minors and directed graph embeddings, by modifying familiar definitions.
Choice Aversion in Directed Networks
TLDR
A recursive logit model is proposed that incorporates choice aversion along the lines introduced by Fudenberg and Strzalecki (2015) and derives optimal flow allocations both sequentially and from a path-choice perspective, which is robust to the presence of overlapping routes.
Directed Graph Minors and Serial-Parallel Width
TLDR
This work proposes new variants of directed graph minors and directed graph embeddings, by modifying familiar definitions, for the class of 2-terminal directed acyclic graphs (TDAGs), where the two definitions coincide, and the class is closed under both operations.
The Tragicomedy of Infrastructure: Exploring Time Inconsistency and Pre-Commitment Strategies in Energy and Transport Infrastructure Investment
This paper blends insights about time inconsistency from two distinct literatures: behavioural economics and macroeconomic policy, and applies them to the problems of infrastructure investment and
Planning to Intervene Under Models of Time Inconsistency
...
...

References

SHOWING 1-10 OF 38 REFERENCES
Computational Issues in Time-Inconsistent Planning
TLDR
A tight upper bound of cost ratio for graphs is shown and confirmed, and the conjecture by Kleinberg and Oren that Akerlof’s structure is indeed the worst case for cost ratio is confirmed, that finding a motivating subgraph is NP-hard, showing that it is generally inefficient to motivate agents by deleting nodes and edges in the graph.
Motivating Time-Inconsistent Agents: A Computational Approach
TLDR
A ( 1 + n ) $(1 + n) -approximation algorithm is devised and proved to prove NP-hardness for ratios greater than n / 3 $\sqrt {n}/3$ and it is argued that the second setting does not permit any efficient approximation unless P = NP.
On the Value of Penalties in Time-Inconsistent Planning
TLDR
This work proposes a penalty-based commitment device that does not delete edges but raises their cost, and proves that optimal penalties are NP-hard to approximate within a ratio of 1.08192.
Procrastination with Variable Present Bias
TLDR
The cost of procrastination is measured as the relative expected cost of the chosen path versus the optimal path, and a novel connection to optimal pricing theory is used to describe the structure of the worst-case task graph for any present-bias distribution.
Planning Problems for Sophisticated Agents with Present Bias
TLDR
A new synthesis of two formalisms --- sophistication and graph-theoretic planning --- uncovers a rich structure that wasn't apparent in the earlier behavioral economics work on this problem, including a range of findings that shed new light on sophisticated behavior.
Doing It Now or Later
Though economists assume that intertemporal preferences are time-consistent, evidence suggests that a person's relative preference for well-being at an earlier moment over a later moment increases as
Planning with Multiple Biases
TLDR
This work proposes a theoretical model for planning with a pair of biases exhibiting an intuitively natural interaction: present bias and sunk-cost bias, and shows how certain natural behavioral phenomena can arise in this model only when agents exhibit both biases.
Procrastination and Obedience
In this lecture I shall focus on situations involving repeated decisions with time inconsistent behavior. Although each choice may be close to maximizing and therefore result in only small losses,
Procrastination, Deadlines, and Performance: Self-Control by Precommitment
TLDR
The answer to three questions is “yes” to the first two questions, and “nO'’ to the third, showing that people have self-control problems, they recognize them, and they try to control them by self-imposing costly deadlines.
...
...