StARLinG Lab
starling-lab
We are an AI Lab interested in making smart machines that humans can use reliably in their lives. Directed by Professor Sriraam Natarajan.
Languages
Top Repositories
BoostSRL: "Boosting for Statistical Relational Learning." A gradient-boosting based approach for learning different types of SRL models.
Development repository for the STARLinG Lab's webpage. Built wtih Jekyll, jQuery, and the minimal-mistakes Jekyll theme.
TensorFlow implementation of "Relational Restricted Boltzmann Machines: A Probabilistic Logic Learning Approach"
Materials for Probabilistic Deep Generative Models Workshop held at IIT Madras, July 3-5 2024.
Knowledge-intensive Gradient Boosting: A unified framework for learning gradient boosted decision trees for regression and classification tasks while leveraging human advice for achieving better performance.
Repositories
19Starling Lab Website
No description provided.
Development repository for the STARLinG Lab's webpage. Built wtih Jekyll, jQuery, and the minimal-mistakes Jekyll theme.
BoostSRL: "Boosting for Statistical Relational Learning." A gradient-boosting based approach for learning different types of SRL models.
Materials for Probabilistic Deep Generative Models Workshop held at IIT Madras, July 3-5 2024.
No description provided.
No description provided.
An implementation of the paper Kokel et al. ICAPS 2021, RePReL: Integrating Relational Planning and Reinforcement Learning for Effective Abstraction.
No description provided.
Knowledge-Intensive learning of Cutset Networks
Knowledge-intensive Gradient Boosting: A unified framework for learning gradient boosted decision trees for regression and classification tasks while leveraging human advice for achieving better performance.
ExSPN: Explaining Sum-Product Networks
Learning Boosted MLN with in-memory Relational Database integration (Malec et al. ILP 2016). This is an extension where wrapper ensures same command line argument structure as MLN-Boost. Most arguments are same as the original MLN-Boost(Khot et al. ICDM 2011) platform. Few that are different have been stated below.
Code repository for the work Relational Boosted Bandits, AAAI'21
This repository contains code base for the slim version of BoostSRL. Performance wise, they are the same, but differs in the volume of redundant code removed in this slim version
TensorFlow implementation of "Relational Restricted Boltzmann Machines: A Probabilistic Logic Learning Approach"
No description provided.
Guided One-shot Concept Induction
The interface lets experts annotate textual data to help a model