Fast Relational Probabilistic Inference and Learning: Approximate Counting via Hypergraphs

Das, Mayukh; Dhami, Devendra Singh; Kunapuli, Gautam; Kersting, Kristian; Natarajan, Sriraam

doi:10.1609/aaai.v33i01.33017816

Cited by 12 publications

(9 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lowd and Domingos (2007) discuss multiple methods for solving the MLE problem including gradient descent, contrastive divergence, diagonal newton, and conjugate gradient. Recent efforts for scaling discriminative MLE weight learning reduce the search space (Farabi et al 2018) by leveraging symmetries to speed up (Ahmadi et al 2012;Van Haaren et al 2015) or approximate the inference subproblem (Sarkhel et al 2016;Das et al 2016Das et al , 2019. For this discussion and for later experiments, we use the diagonal Newton method (DN ) as the representative MLE weight learner as it was found to be the among the most effective for MLE weight learning in MLNs (Lowd and Domingos 2007) and is the default optimizer for Tuffy (Niu et al 2011).…”

Section: Maximum-likelihood Estimationmentioning

confidence: 99%

“…Typically, the weights of the rules are learned through maximizing some form of likelihood function (Bach et al 2013;Lowd and Domingos 2007;Singla and Domingos 2005;Kok and Domingos 2005;Chou et al 2016;Sarkhel et al 2016;Das et al 2016Das et al , 2019Farabi et al 2018). This is a well-motivated approach if the downstream objective makes use of the probability density function directly.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A taxonomy of weight learning methods for statistical relational learning

et al. 2021

View full text Add to dashboard Cite

Statistical relational learning (SRL) frameworks are effective at defining probabilistic models over complex relational data. They often use weighted first-order logical rules where the weights of the rules govern probabilistic interactions and are usually learned from data. Existing weight learning approaches typically attempt to learn a set of weights that maximizes some function of data likelihood; however, this does not always translate to optimal performance on a desired domain metric, such as accuracy or F1 score. In this paper, we introduce a taxonomy of search-based weight learning approaches for SRL frameworks that directly optimize weights on a chosen domain performance metric. To effectively apply these search-based approaches, we introduce a novel projection, referred to as scaled space (SS), that is an accurate representation of the true weight space. We show that SS removes redundancies in the weight space and captures the semantic distance between the possible weight configurations. In order to improve the efficiency of search, we also introduce an approximation of SS which simplifies the process of sampling weight configurations. We demonstrate these approaches on two state-of-the-art SRL frameworks: Markov logic networks and probabilistic soft logic. We perform empirical evaluation on five real-world datasets and evaluate them each on two different metrics. We also compare them against four other weight learning approaches. Our experimental results show that our proposed search-based approaches outperform likelihood-based approaches and yield up to a 10% improvement across a variety of performance metrics. Further, we perform an extensive evaluation to measure the robustness of our approach to different initializations and hyperparameters. The results indicate that our approach is both accurate and robust.

show abstract

Section: Maximum-likelihood Estimationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A taxonomy of weight learning methods for statistical relational learning

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The experimental results obtained showed that the proposed method have better performance when the networks are heterophilous. Das et al [8] proposed fast relational probabilistic inference and learning by using approximate counting via hypergraphs. Their experimental results showed that the efficiency of these approximations allows these models to perform significantly faster than state-of-theart, which can be successfully applied to several complex statistical relational models, without sacrificing effectiveness.…”

Section: Application Of Srlmentioning

confidence: 99%

Statistical Relational Learning: A State-of-the-Art Review

Kastrati¹,

Biba²

2019

Journal of Engineering Technology and Applied Sciences

View full text Add to dashboard Cite

The objective of this paper is to review the state-of-the-art of statistical relational learning (SRL) models developed to deal with machine learning and data mining in relational domains in presence of missing, partially observed, and/or noisy data. It starts by giving a general overview of conventional graphical models, first-order logic and inductive logic programming approaches as needed for background. The historical development of each SRL key model is critically reviewed. The study also focuses on the practical application of SRL techniques to a broad variety of areas and their limitations.

show abstract

“…Algorithm 1 outlines the key steps involved in our approach. KCLN() is the main procedure [lines: [1][2][3][4][5][6][7][8][9][10][11][12][13][14] that trains a Column Network using both the data (the knowledge graph G) and the human advice (set of preference rules P). It returns a K-CLN C P where θ P are the network parameters, which are initialized to any arbitrary value (0 in our case; [line: 3]).…”

Section: The Algorithmmentioning

confidence: 99%

Human-Guided Learning of Column Networks: Augmenting Deep Learning with Advice

Das¹,

Yu²,

Dhami³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

Recently, deep models have been successfully applied in several applications, especially with low-level representations. However, sparse, noisy samples and structured domains (with multiple objects and interactions) are some of the open challenges in most deep models. Column Networks, a deep architecture, can succinctly capture such domain structure and interactions, but may still be prone to sub-optimal learning from sparse and noisy samples. Inspired by the success of human-advice guided learning in AI, especially in data-scarce domains, we propose Knowledge-augmented Column Networks that leverage human advice/knowledge for better learning with noisy/sparse samples. Our experiments demonstrate that our approach leads to either superior overall performance or faster convergence (i.e., both effective and efficient).

show abstract

Fast Relational Probabilistic Inference and Learning: Approximate Counting via Hypergraphs

Cited by 12 publications

References 23 publications

A taxonomy of weight learning methods for statistical relational learning

A taxonomy of weight learning methods for statistical relational learning

Statistical Relational Learning: A State-of-the-Art Review

Human-Guided Learning of Column Networks: Augmenting Deep Learning with Advice

Contact Info

Product

Resources

About