Skip to search formSkip to main contentSkip to account menu
RePReL: Integrating Relational Planning and Reinforcement Learning for Effective Abstraction
TLDR
The experiments clearly show that RePReL framework not only achieves better per- formance and efficient learning on the task at hand but also demonstrates better generalization to unseen tasks.
View Paper
Graph Sparsification via Meta-Learning
TLDR
A novel graph sparsification approach for semisupervised learning on undirected attributed graphs using meta-gradients to solve the optimization problem, essentially treating the graph adjacency matrix as hyperparameter to optimize.
A Unified Framework for Knowledge Intensive Gradient Boosting: Leveraging Human Experts for Noisy Sparse Domains
TLDR
A unified framework for both classification and regression settings that can both effectively and efficiently incorporate qualitative constraints to accelerate learning to a better model is developed.
View via Publisher
Morpheme Extraction Task at FIRE 2012--2013
TLDR
The Morpheme Extraction task (MET) was organized for the first time in FIRE 2012 and subsequently in FIRE 2013 and the goals, data, tasks, participants, evaluation and results obtained are described.
View on ACM
A Probabilistic Approach to Extract Qualitative Knowledge for Early Prediction of Gestational Diabetes
TLDR
This work applies the Qualitative Knowledge Extraction method toward early prediction of gestational diabetes on clinical study data and empirical results demonstrate that the extracted rules are both interpretable and valid.
View on PubMed
Action Space Reduction for Planning Domains
TLDR
An automated way of reducing the action spaces of Reinforcement Learning environments, by leveraging lifted mutex groups is proposed, showing a significant reduction in the action space size of the RL environments.
Human-guided Collaborative Problem Solving: A Natural Language based Framework
TLDR
The ability of this framework to address the key challenges of collaborative problem solving is illustrated by demonstrat-ing it on a collaborative building task in a Minecraft-based blocksworld domain.
View PDF on arXiv
Dynamic probabilistic logic models for effective task-specific abstractions in RL
TLDR
The dynamic probabilistic logic model is adapted to specify bisimilarity conditions of the MDPs to obtain safe and effective task-specific state abstractions and empirical evaluations show that such abstractions can result in efficient learning and effective transfer.
Hybrid Deep RePReL: Integrating Relational Planning and Reinforcement Learning for Information Fusion
Fusion of high-level symbolic reasoning with lower level signal-based reasoning has attracted significant attention. We propose an architecture that integrates the high-level symbolic domain…
View on IEEE
How to Reduce Action Space for Planning Domains? (Student Abstract)
TLDR
This paper presents a method that reduces the parameters of operators in AI planning domains by introducing a parameter seed set problem and casting it as a classical planning task and shows that the proposed method significantly reduces the number of actions in the RL environments originating fromAI planning domains.
View via Publisher
...
...
By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy, Terms of Service, and Dataset License