Improving Multi-Step Prediction of Learned Time Series Models

arXiv 2015 35.9 method

TLDR

A method improving multi-step time series prediction using imitation learning with theoretical guarantees, tested on dynamic systems and video textures.

Reasoning

The paper presents a novel reduction of multi-step prediction to imitation learning, providing a theoretical guarantee and empirical improvements in two domains. However, the scope is limited to time series and does not explicitly address world models or simulation, which may reduce relevance to those keywords.

Read-first score

Read-first score 35.9, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 8.

Citation impact 18%

98.7

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.98670776

Citation velocity 12%

58.3

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=16.50

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=experiment,result

Recency 6%

20.8

Uses a gentle age decay so recent papers surface without erasing older foundations. 2015

Topical relevance 29%

11.4

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=False; code=False; dataset=False; markers=none

Field roles

Foundation

Rank sensitivity

Stability: volatile; rank range: 387.

Keyword Scores

world dynamics prediction

video world model

world model

world simulator

generative world model

interactive world model

model-based reinforcement learning world model