scholarly journals A taxonomy of weight learning methods for statistical relational learning

2021 ◽  
Author(s):  
Sriram Srinivasan ◽  
Charles Dickens ◽  
Eriq Augustine ◽  
Golnoosh Farnadi ◽  
Lise Getoor
Keyword(s):  
Probabilistic Models ◽  
Performance Metrics ◽  
Relational Learning ◽  
Empirical Evaluation ◽  
Statistical Relational Learning ◽  
Weight Space ◽  
Learning Approaches ◽  
Extensive Evaluation ◽  
Real World Datasets ◽  
Weight Learning

AbstractStatistical 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.

scholarly journals A comparison of statistical relational learning and graph neural networks for aggregate graph queries

2021 ◽  
Author(s):  
Varun Embar ◽  
Sriram Srinivasan ◽  
Lise Getoor
Keyword(s):  
Neural Networks ◽  
Graph Mining ◽  
Probabilistic Models ◽  
Relational Learning ◽  
Empirical Evaluation ◽  
Statistical Relational Learning ◽  
Average Error ◽  
Graph Queries ◽  
Graph Neural Networks ◽  
Node Labels

AbstractStatistical relational learning (SRL) and graph neural networks (GNNs) are two powerful approaches for learning and inference over graphs. Typically, they are evaluated in terms of simple metrics such as accuracy over individual node labels. Complex aggregate graph queries (AGQ) involving multiple nodes, edges, and labels are common in the graph mining community and are used to estimate important network properties such as social cohesion and influence. While graph mining algorithms support AGQs, they typically do not take into account uncertainty, or when they do, make simplifying assumptions and do not build full probabilistic models. In this paper, we examine the performance of SRL and GNNs on AGQs over graphs with partially observed node labels. We show that, not surprisingly, inferring the unobserved node labels as a first step and then evaluating the queries on the fully observed graph can lead to sub-optimal estimates, and that a better approach is to compute these queries as an expectation under the joint distribution. We propose a sampling framework to tractably compute the expected values of AGQs. Motivated by the analysis of subgroup cohesion in social networks, we propose a suite of AGQs that estimate the community structure in graphs. In our empirical evaluation, we show that by estimating these queries as an expectation, SRL-based approaches yield up to a 50-fold reduction in average error when compared to existing GNN-based approaches.

scholarly journals Induction of Interpretable Possibilistic Logic Theories from Relational Data

2017 ◽  
Author(s):  
Ondrej Kuzelka ◽  
Jesse Davis ◽  
Steven Schockaert
Keyword(s):  
Neural Networks ◽  
Probabilistic Models ◽  
Relational Learning ◽  
Statistical Relational Learning ◽  
Relational Data ◽  
Markov Logic Networks ◽  
Relational Models ◽  
Markov Logic ◽  
Possibilistic Logic ◽  
Interpretable Models

The field of statistical relational learning (SRL) is concerned with learning probabilistic models from relational data. Learned SRL models are typically represented using some kind of weighted logical formulas, which makes them considerably more interpretable than those obtained by e.g. neural networks. In practice, however, these models are often still difficult to interpret correctly, as they can contain many formulas that interact in non-trivial ways and weights do not always have an intuitive meaning. To address this, we propose a new SRL method which uses possibilistic logic to encode relational models. Learned models are then essentially stratified classical theories, which explicitly encode what can be derived with a given level of certainty. Compared to Markov Logic Networks (MLNs), our method is faster and produces considerably more interpretable models.

scholarly journals BOWL: Bayesian Optimization for Weight Learning in Probabilistic Soft Logic

2020 ◽  
Vol 34 (06) ◽  
pp. 10267-10275
Author(s):  
Sriram Srinivasan ◽  
Golnoosh Farnadi ◽  
Lise Getoor
Keyword(s):  
Performance Metrics ◽  
Gaussian Process Regression ◽  
Semantic Distance ◽  
Bayesian Optimization ◽  
Learning Approaches ◽  
Learning Framework ◽  
Performance Metric ◽  
Soft Logic ◽  
Weight Learning ◽  
Domain Performance

Probabilistic soft logic (PSL) is a statistical relational learning framework that represents complex relational models with weighted first-order logical rules. The weights of the rules in PSL indicate their importance in the model and influence the effectiveness of the model on a given task. Existing weight learning approaches often 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 new weight learning approach called Bayesian optimization for weight learning (BOWL) based on Gaussian process regression that directly optimizes weights on a chosen domain performance metric. The key to the success of our approach is a novel projection that captures the semantic distance between the possible weight configurations. Our experimental results show that our proposed approach outperforms likelihood-based approaches and yields up to a 10% improvement across a variety of performance metrics. Further, we performed experiments to measure the scalability and robustness of our approach on various realworld datasets.

scholarly journals A Comparative Study of Distributional and Symbolic Paradigms for Relational Learning

2019 ◽  
Author(s):  
Sebastijan Dumancic ◽  
Alberto Garcia-Duran ◽  
Mathias Niepert
Keyword(s):  
Euclidean Space ◽  
Knowledge Base ◽  
Relational Learning ◽  
Statistical Relational Learning ◽  
Representation Learning ◽  
The Other ◽  
Learning Approaches ◽  
Relational Knowledge ◽  
Knowledge Graphs ◽  
Relational Classification

Many real-world domains can be expressed as graphs and, more generally, as multi-relational knowledge graphs. Though reasoning and learning with knowledge graphs has traditionally been addressed by symbolic approaches such as Statistical relational learning, recent methods in (deep) representation learning have shown promising results for specialised tasks such as knowledge base completion. These approaches, also known as distributional, abandon the traditional symbolic paradigm by replacing symbols with vectors in Euclidean space. With few exceptions, symbolic and distributional approaches are explored in different communities and little is known about their respective strengths and weaknesses. In this work, we compare distributional and symbolic relational learning approaches on various standard relational classification and knowledge base completion tasks. Furthermore, we analyse the properties of the datasets and relate them to the performance of the methods in the comparison. The results reveal possible indicators that could help in choosing one approach over the other for particular knowledge graphs.

scholarly journals Machine Learning Methods Applied to the Prediction of Pseudo-nitzschia spp. Blooms in the Galician Rias Baixas (NW Spain)

2021 ◽  
Vol 10 (4) ◽  
pp. 199
Author(s):  
Francisco M. Bellas Aláez ◽  
Jesus M. Torres Palenzuela ◽  
Evangelos Spyrakos ◽  
Luis González Vilas
Keyword(s):  
Machine Learning ◽  
Performance Metrics ◽  
Prediction Models ◽  
Support Vector ◽  
False Alarms ◽  
Learning Approaches ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Rías Baixas ◽  
New Algorithms

This work presents new prediction models based on recent developments in machine learning methods, such as Random Forest (RF) and AdaBoost, and compares them with more classical approaches, i.e., support vector machines (SVMs) and neural networks (NNs). The models predict Pseudo-nitzschia spp. blooms in the Galician Rias Baixas. This work builds on a previous study by the authors (doi.org/10.1016/j.pocean.2014.03.003) but uses an extended database (from 2002 to 2012) and new algorithms. Our results show that RF and AdaBoost provide better prediction results compared to SVMs and NNs, as they show improved performance metrics and a better balance between sensitivity and specificity. Classical machine learning approaches show higher sensitivities, but at a cost of lower specificity and higher percentages of false alarms (lower precision). These results seem to indicate a greater adaptation of new algorithms (RF and AdaBoost) to unbalanced datasets. Our models could be operationally implemented to establish a short-term prediction system.

scholarly journals An empirical evaluation of cost-based federated SPARQL query processing engines

Semantic Web ◽  
2021 ◽  
pp. 1-26
Author(s):  
Umair Qudus ◽  
Muhammad Saleem ◽  
Axel-Cyrille Ngonga Ngomo ◽  
Young-Koo Lee
Keyword(s):  
Query Processing ◽  
Detailed Analysis ◽  
Performance Metrics ◽  
Empirical Evaluation ◽  
Sparql Query ◽  
Evaluation Metrics ◽  
Future Cost ◽  
Query Plan ◽  
Fine Grained ◽  
Runtime Performance

Finding a good query plan is key to the optimization of query runtime. This holds in particular for cost-based federation engines, which make use of cardinality estimations to achieve this goal. A number of studies compare SPARQL federation engines across different performance metrics, including query runtime, result set completeness and correctness, number of sources selected and number of requests sent. Albeit informative, these metrics are generic and unable to quantify and evaluate the accuracy of the cardinality estimators of cost-based federation engines. To thoroughly evaluate cost-based federation engines, the effect of estimated cardinality errors on the overall query runtime performance must be measured. In this paper, we address this challenge by presenting novel evaluation metrics targeted at a fine-grained benchmarking of cost-based federated SPARQL query engines. We evaluate five cost-based federated SPARQL query engines using existing as well as novel evaluation metrics by using LargeRDFBench queries. Our results provide a detailed analysis of the experimental outcomes that reveal novel insights, useful for the development of future cost-based federated SPARQL query processing engines.

scholarly journals A Comprehensive Study of Artificial Intelligence and Machine Learning Approaches in Confronting the Coronavirus (COVID-19) Pandemic

2021 ◽  
pp. 002073142110174
Author(s):  
Md Mijanur Rahman ◽  
Fatema Khatun ◽  
Ashik Uzzaman ◽  
Sadia Islam Sami ◽  
Md Al-Amin Bhuiyan ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Health Care ◽  
Language Processing ◽  
Probabilistic Models ◽  
Health Care Systems ◽  
Learning Approaches ◽  
Care Systems ◽  
Novel Coronavirus ◽  
Comprehensive Study

The novel coronavirus disease (COVID-19) has spread over 219 countries of the globe as a pandemic, creating alarming impacts on health care, socioeconomic environments, and international relationships. The principal objective of the study is to provide the current technological aspects of artificial intelligence (AI) and other relevant technologies and their implications for confronting COVID-19 and preventing the pandemic’s dreadful effects. This article presents AI approaches that have significant contributions in the fields of health care, then highlights and categorizes their applications in confronting COVID-19, such as detection and diagnosis, data analysis and treatment procedures, research and drug development, social control and services, and the prediction of outbreaks. The study addresses the link between the technologies and the epidemics as well as the potential impacts of technology in health care with the introduction of machine learning and natural language processing tools. It is expected that this comprehensive study will support researchers in modeling health care systems and drive further studies in advanced technologies. Finally, we propose future directions in research and conclude that persuasive AI strategies, probabilistic models, and supervised learning are required to tackle future pandemic challenges.

scholarly journals Competitive Caching with Machine Learned Advice

2021 ◽  
Vol 68 (4) ◽  
pp. 1-25
Author(s):  
Thodoris Lykouris ◽  
Sergei Vassilvitskii
Keyword(s):  
Online Algorithms ◽  
Empirical Evaluation ◽  
Optimal Solution ◽  
Poor Performance ◽  
Machine Learning Algorithms ◽  
Average Error ◽  
Generalization Error ◽  
Worst Case ◽  
Future Events ◽  
Real World Datasets

Traditional online algorithms encapsulate decision making under uncertainty, and give ways to hedge against all possible future events, while guaranteeing a nearly optimal solution, as compared to an offline optimum. On the other hand, machine learning algorithms are in the business of extrapolating patterns found in the data to predict the future, and usually come with strong guarantees on the expected generalization error. In this work, we develop a framework for augmenting online algorithms with a machine learned predictor to achieve competitive ratios that provably improve upon unconditional worst-case lower bounds when the predictor has low error. Our approach treats the predictor as a complete black box and is not dependent on its inner workings or the exact distribution of its errors. We apply this framework to the traditional caching problem—creating an eviction strategy for a cache of size k . We demonstrate that naively following the oracle’s recommendations may lead to very poor performance, even when the average error is quite low. Instead, we show how to modify the Marker algorithm to take into account the predictions and prove that this combined approach achieves a competitive ratio that both (i) decreases as the predictor’s error decreases and (ii) is always capped by O (log k ), which can be achieved without any assistance from the predictor. We complement our results with an empirical evaluation of our algorithm on real-world datasets and show that it performs well empirically even when using simple off-the-shelf predictions.

scholarly journals IoT Botnet Anomaly Detection Using Unsupervised Deep Learning

Electronics ◽  
2021 ◽  
Vol 10 (16) ◽  
pp. 1876
Author(s):  
Ioana Apostol ◽  
Marius Preda ◽  
Constantin Nila ◽  
Ion Bica
Keyword(s):  
Deep Learning ◽  
Detection System ◽  
False Positive Rate ◽  
Empirical Evaluation ◽  
Learning Approaches ◽  
Promising Alternative ◽  
Positive Rate ◽  
Unsupervised Deep Learning ◽  
Attack Surface ◽  
Iot Devices

The Internet of Things has become a cutting-edge technology that is continuously evolving in size, connectivity, and applicability. This ecosystem makes its presence felt in every aspect of our lives, along with all other emerging technologies. Unfortunately, despite the significant benefits brought by the IoT, the increased attack surface built upon it has become more critical than ever. Devices have limited resources and are not typically created with security features. Lately, a trend of botnet threats transitioning to the IoT environment has been observed, and an army of infected IoT devices can expand quickly and be used for effective attacks. Therefore, identifying proper solutions for securing IoT systems is currently an important and challenging research topic. Machine learning-based approaches are a promising alternative, allowing the identification of abnormal behaviors and the detection of attacks. This paper proposes an anomaly-based detection solution that uses unsupervised deep learning techniques to identify IoT botnet activities. An empirical evaluation of the proposed method is conducted on both balanced and unbalanced datasets to assess its threat detection capability. False-positive rate reduction and its impact on the detection system are also analyzed. Furthermore, a comparison with other unsupervised learning approaches is included. The experimental results reveal the performance of the proposed detection method.

Empirical Evaluation of Generic Weighted Cubicle Pattern and LBP Derivatives for Abnormality Detection in Mammogram Images

2015 ◽  
Vol 15 (01) ◽  
pp. 1550001 ◽  
Author(s):  
A. Suruliandi ◽  
G. Murugeswari ◽  
P. Arockia Jansi Rani
Keyword(s):  
Image Segmentation ◽  
Performance Metrics ◽  
Empirical Evaluation ◽  
Rate Sensitivity ◽  
Digital Mammogram ◽  
Abnormality Detection ◽  
Texture Descriptor ◽  
Texture Pattern ◽  
Local Texture ◽  
Mammogram Images

Digital image processing techniques are very useful in abnormality detection in digital mammogram images. Nowadays, texture-based image segmentation of digital mammogram images is very popular due to its better accuracy and precision. Local binary pattern (LBP) descriptor has attracted many researchers working in the field of texture analysis of digital images. Because of its success, many texture descriptors have been introduced as variants of LBP. In this work, we propose a novel texture descriptor called generic weighted cubicle pattern (GWCP) and we analyzed the proposed operator for texture image classification. We also performed abnormality detection through mammogram image segmentation using k-Nearest Neighbors (KNN) algorithm and compared the performance of the proposed texture descriptor with LBP and other variants of LBP namely local ternary pattern (LTPT), extended local texture pattern (ELTP) and local texture pattern (LTPS). For evaluation, we used the performance metrics such as accuracy, error rate, sensitivity, specificity, under estimation fraction and over estimation fraction. The results prove that the proposed method outperforms other descriptors in terms of abnormality detection in mammogram images.

Sign in / Sign up

Export Citation Format

Share Document