Combining Domain Knowledge Research Articles

A Bayesian network approach for predicting seismic liquefaction based on interpretive structural modeling

ABSTRACTThe Bayesian network (BN) is a type of graphical network based on probabilistic inference that has been gradually applied to assessment of seismic liquefaction potential. However, how to construct a robust BN remains underexplored in this field. This paper aims to present an efficient hybrid approach combining domain knowledge and data to construct a BN that facilitates the integration of multiple factors and the quantification of uncertainties within a network model to assess seismic liquefaction. Initially, only using given domain knowledge, a naive network model can be constructed using interpretive structural modeling. Thereafter, some effective information about the naive model is provided to construct a robust model using structural learning of BN from historic data. Finally, the returning predictive results and the predictive results are compared to other methods including non-probabilistic and probabilistic models for seismic liquefaction using the metrics of the overall accuracy, the area under the curve of receiver operating characteristic, prediction, recall and F1 score. The methodology proposed in this paper achieved better performance, and we discussed the power and value of the proposed approach at the end of this paper, which suggest that BN is a good alternative tool for seismic liquefaction prediction.

Combining domain knowledge and machine learning for robust fall detection

AbstractThis paper presents a method for combining domain knowledge and machine learning (CDKML) for classifier generation and online adaptation. The method exploits advantages in domain knowledge and machine learning as complementary information sources. Whereas machine learning may discover patterns in interest domains that are too subtle for humans to detect, domain knowledge may contain information on a domain not present in the available domain dataset. CDKML has three steps. First, prior domain knowledge is enriched with relevant patterns obtained by machine learning to create an initial classifier. Second, genetic algorithms refine the classifier. Third, the classifier is adapted online on the basis of user feedback using the Markov decision process. CDKML was applied in fall detection. Tests showed that the classifiers developed by CDKML have better performance than machine‐learning classifiers generated on a training dataset that does not adequately represent all real‐life cases of the learned concept. The accuracy of the initial classifier was 10 percentage points higher than the best machine‐learning classifier and the refinement added 3 percentage points. The online adaptation improved the accuracy of the refined classifier by an additional 15 percentage points.

Three-layer Activity Recognition Combining Domain Knowledge and Meta-classification

One of the essential tasks of healthcare and smart-living systems is to recognize the current activity of a particular user. Such activity recognition (AR) is demanding when only limited sensors are used, such as accelerometers. Given a small number of accelerometers, intelligent AR systems often use simple architectures, either general or specific for their AR. In this paper, a system for AR named TriLAR is presented. TriLAR has an AR-specific architecture consisting of three layers: (i) a bottom layer, where an arbitrary number of AR methods can be used to recognize the current activity; (ii) a middle layer, where the predictions from the bottom-layer methods are inputs for a hierarchical structure that combines domain knowledge and meta-classification; and (iii) a top layer, where a hidden Markov model is used to correct spurious transitions between the recognized activities from the middle layer. The middle layer has a hierarchical, three-level structure. First, a meta-classifier is used to make the initial separation between the most distinct activities. Second, domain knowledge in the form of rules is used to differentiate between the remaining activities, recognizing those of interest (i.e., static activities). Third, another meta-classifier deals with the remaining activities. In this way, each activity is recognized by the method best suited to it, leaving unrecognized activities to the next method. This architecture was tested on a dataset recorded using ten volunteers who acted out a complex, real-life scenario while wearing accelerometers placed on the chest, thigh, and ankle. The results show that TriLAR successfully recognized elementary activities using one or two sensors and significantly outperformed three standard, single-layer methods with all sensor placements.

SEWEBAR-CMS: semantic analytical report authoring for data mining results

SEWEBAR-CMS is a set of extensions for the Joomla! Content Management System (CMS) that extends it with functionality required to serve as a communication platform between the data analyst, domain expert and the report user. SEWEBAR-CMS integrates with existing data mining software through PMML. Background knowledge is entered via a web-based elicitation interface and is preserved in documents conforming to the proposed Background Knowledge Exchange Format (BKEF) specification. SEWEBAR-CMS offers web service integration with semantic knowledge bases, into which PMML and BKEF data are stored. Combining domain knowledge and mining model visualizations with results of queries against the knowledge base, the data analyst conveys the results of the mining through a semi-automatically generated textual analytical report to the end user. The paper demonstrates the use of SEWEBAR-CMS on a real-world task from the cardiological domain and presents a user study showing that the proposed report authoring support leads to a statistically significant decrease in the time needed to author the analytical report.

A Primer on Learning in Bayesian Networks for Computational Biology

Bayesian networks (BNs) provide a neat and compact representation for expressing joint probability distributions (JPDs) and for inference. They are becoming increasingly important in the biological sciences for the tasks of inferring cellular networks [1], modelling protein signalling pathways [2], systems biology, data integration [3], classification [4], and genetic data analysis [5]. The representation and use of probability theory makes BNs suitable for combining domain knowledge and data, expressing causal relationships, avoiding overfitting a model to training data, and learning from incomplete datasets. The probabilistic formalism provides a natural treatment for the stochastic nature of biological systems and measurements. This primer aims to introduce BNs to the computational biologist, focusing on the concepts behind methods for learning the parameters and structure of models, at a time when they are becoming the machine learning method of choice. There are many applications in biology where we wish to classify data; for example, gene function prediction. To solve such problems, a set of rules are required that can be used for prediction, but often such knowledge is unavailable, or in practice there turn out to be many exceptions to the rules or so many rules that this approach produces poor results. Machine learning approaches often produce better results, where a large number of examples (the training set) is used to adapt the parameters of a model that can then be used for performing predictions or classifications on data. There are many different types of models that may be required and many different approaches to training the models, each with its pros and cons. An excellent overview of the topic can be found in [6] and [7]. Neural networks, for example, are often able to learn a model from training data, but it is often difficult to extract information about the model, which with other methods can provide valuable insights into the data or problem being solved. A common problem in machine learning is overfitting, where the learned model is too complex and generalises poorly to unseen data. Increasing the size of the training dataset may reduce this; however, this assumes more training data is readily available, which is often not the case. In addition, often it is important to determine the uncertainty in the learned model parameters or even in the choice of model. This primer focuses on the use of BNs, which offer a solution to these issues. The use of Bayesian probability theory provides mechanisms for describing uncertainty and for adapting the number of parameters to the size of the data. Using a graphical representation provides a simple way to visualise the structure of a model. Inspection of models can provide valuable insights into the properties of the data and allow new models to be produced.

Using Bayesian belief networks to analyse the stochastic dependence between interevent time and size of earthquakes

The purpose of this article is to show how Bayesian belief networks can beused in analysis of the sequence of the earthquakes which have occurred in a region, to study the interaction among the variables characterizing eachevent. These relationships can be represented by means of graphs consistingof vertices and edges; the vertices correspond to random variables, whilethe edges express properties of conditional independence. We have examinedItalian seismicity as reported in two data bases, the NT4.1.1 catalogue and the ZS.4 zonation, and taken into account three variables: the size of thequake, the time elapsed since the previous event, and the time before the subsequent one. Assigning different independence relationships among these variables, first two couples of bivariate models, and then eight trivariatemodels have been defined. After presenting the main elements constituting a Bayesian belief network, we introduce the principal methodological aspects concerning estimation and model comparison. Following a fully Bayesian approach, prior distributions are assigned on both parameters and structuresby combining domain knowledge and available information on homogeneous seismogenic zones. Two case studies are used to illustrate in detail the procedure followed to evaluate the fitting of each model to the data sets andcompare the performance of alternative models. All eighty Italian seismogenic zones have been analysed in the same way; the results obtained are reportedbriefly. We also show how to account for model uncertainty in predicting a quantity of interest, such as the time of the next event.