Rule-based Classification Research Articles

A Parcel-Based Deep-Learning Classification to Map Local Climate Zones From Sentinel-2 Images

Local climate zones (LCZ) describe urban surface structures, supporting studies of urban heat islands, sustainable urbanization, and energy balance. The existing studies mapped LCZs from satellite images using scene-based classification, which trained deep-learning classifiers by labeled image patches, segmented satellite images into patches by sliding windows to match the size of training data, and finally classified the segmented patches to obtain LCZ maps. However, sliding windows are different from the real footprints of LCZs, which leads to large errors in classification. To address this problem, this article proposes a parcel-based method for LCZ classification using Sentinel-2 images, road networks, and elevation data. First, the Sentinel-2 images are segmented by the road network to obtain the land parcels as classification units. Second, each image parcel is standardized to match the training dataset, So2Sat LCZ42. Third, the trained convolutional neural network (CNN) is used to classify the standardized parcels into LCZs. Finally, the building height information derived from elevation data is used to refine the LCZs by a rule-based classifier. The results of the four test sites show that the overall accuracy of our method is 0.75, higher than the sliding-window-based method's accuracy of 0.47. Additional simulation experiments demonstrated that parcels derived from road networks can reduce the mixture effect in image patches, and parcel standardization can ensure the transferability of the CNN model trained by regular image patches. Considering that the road network and elevation data are widely available, the proposed method has the potential of mapping LCZs in large areas.

Predicting heart failure using data mining with Rough set theory and Fuzzy Petri Net

The Rough Set Theory (RST) is a method that has proven its efficiency and simplicity in machine learning and successfully developing now a day’s vastly and rapidly. Fuzzy Petri nets (FPNs) are a potential modelling technique, which is used for knowledge representation and reasoning of rule-based expert systems. Though, RST has the efficiency in dimension reduction, it has to be proved with evidence by creating model using FPN with rule based reasoning. In this paper the induction of decision rules by RST executed with Fuzzy Petri Nets (FPN) is analyzed in the sense how it performed better than other data mining classifiers. The rule-based classifiers like jrip, part R, zero R are used for the comparison purposes. The knowledge captured from the rules through the best reduct has performed with the efficiency of RST and formulating rules by FPN. This paper experiments the Heart failure data to investigate the decision making from the rules generated by LERS system of RST with the approach of FPN. The heart failures during the follow up period of the patents are predicted with the pattern recognized form the data using the above process and the best evaluators are found during the experiments were pictured.

Semi-automatic multi-segmentation classification for land cover change dynamics in North Macedonia from 1988 to 2014

Land cover assessment and monitoring are essential for sustainable management of natural resources and environmental protection. Object-based image analysis (OBIA) for land cover classification has become an area of interest due to its superiority over the pixel-based classification method. The main objective of this paper is developing a method for land cover classification on the national and sub-national level in the Republic of North Macedonia for mapping and monitoring the land cover changes in the study area from 1988 to 2014. For that purpose, in this study, we combine OBIA with rule set semi-automated multi-segmentation classification for large-scale areas over medium-resolution satellite imagery. Thus, Landsat image collections over North Macedonia have been combined with topographic and settlement layers for land cover classification. Based on the knowledge of certain land cover features, rule-based classification has been developed using two different segmentation parameters. The results show that the overall agreement of the new semi-automatic classification method developed for North Macedonia is 83%. The most significant change in the land cover can be noticed in the forest class, with a total increase of 8% on national and 15% in the South-East region. These results confirm that this new semi-automatic, cost-effective, and accurate land cover classification method can be easily employed and adjusted for different study areas and can be used in numerous remote sensing applications.

ReG-Rules: An Explainable Rule-Based Ensemble Learner for Classification

The learning of classification models to predict class labels of new and previously unseen data instances is one of the most essential tasks in data mining. A popular approach to classification is ensemble learning, where a combination of several diverse and independent classification models is used to predict class labels. Ensemble models are important as they tend to improve the average classification accuracy over any member of the ensemble. However, classification models are also often required to be explainable to reduce the risk of irreversible wrong classification. Explainability of classification models is needed in many critical applications such as stock market analysis, credit risk evaluation, intrusion detection, etc. Unfortunately, ensemble learning decreases the level of explainability of the classification, as the analyst would have to examine many decision models to gain insights about the causality of the prediction. The aim of the research presented in this paper is to create an ensemble method that is explainable in the sense that it presents the human analyst with a conditioned view of the most relevant model aspects involved in the prediction. To achieve this aim the authors developed a rule-based explainable ensemble classifier termed Ranked ensemble G-Rules (ReG-Rules) which gives the analyst an extract of the most relevant classification rules for each individual prediction. During the evaluation process ReG-Rules was evaluated in terms of its theoretical computational complexity, empirically on benchmark datasets and qualitatively with respect to the complexity and readability of the induced rule sets. The results show that ReG-Rules scales linearly, delivers a high accuracy and at the same time delivers a compact and manageable set of rules describing the predictions made.

Intrusion detection system using a new fuzzy rule-based classification system based on genetic algorithm

Intrusion can compromise the integrity, confidentiality, or availability of a computer system. Intrusion Detection System (IDS) is a type of security software designed to monitor network traffic and identify network intrusions. In this paper, A Fuzzy Rule – Based classification system is used to detect intrusion in a computer network. In order to improve the classification rate, a new method is proposed based on Genetic Algorithm (GA) for rule weights specification. The proposed method is tested on KDD99 dataset. Experimental results show the proposed method improves the performance of the fuzzy rule-based classification systems in terms of detection rate and false alarm rate.

Comparison of segmentation-free and segmentation-dependent computer-aided diagnosis of breast masses on a public mammography dataset

PurposeTo compare machine learning methods for classifying mass lesions on mammography images that use predefined image features computed over lesion segmentations to those that leverage segmentation-free representation learning on a standard, public evaluation dataset. MethodsWe apply several classification algorithms to the public Curated Breast Imaging Subset of the Digital Database for Screening Mammography (CBIS-DDSM), in which each image contains a mass lesion. Segmentation-free representation learning techniques for classifying lesions as benign or malignant include both a Bag-of-Visual-Words (BoVW) method and a Convolutional Neural Network (CNN). We compare classification performance of these techniques to that obtained using two different segmentation-dependent approaches from the literature that rely on specific combinations of end classifiers (e.g. linear discriminant analysis, neural networks) and predefined features computed over the lesion segmentation (e.g. spiculation measure, morphological characteristics, intensity metrics). ResultsWe report area under the receiver operating characteristic curve (AZ) values for malignancy classification on CBIS-DDSM for each technique. We find average AZ values of 0.73 for a segmentation-free BoVW method, 0.86 for a segmentation-free CNN method, 0.75 for a segmentation-dependent linear discriminant analysis of Rubber-Band Straightening Transform features, and 0.58 for a hybrid rule-based neural network classification using a small number of hand-designed features. ConclusionsWe find that malignancy classification performance on the CBIS-DDSM dataset using segmentation-free BoVW features is comparable to that of the best segmentation-dependent methods we study, but also observe that a common segmentation-free CNN model substantially and significantly outperforms each of these (p < 0.05). These results reinforce recent findings suggesting that representation learning techniques such as BoVW and CNNs are advantageous for mammogram analysis because they do not require lesion segmentation, the quality and specific characteristics of which can vary substantially across datasets. We further observe that segmentation-dependent methods achieve performance levels on CBIS-DDSM inferior to those achieved on the original evaluation datasets reported in the literature. Each of these findings reinforces the need for standardization of datasets, segmentation techniques, and model implementations in performance assessments of automated classifiers for medical imaging.

Hybrid bat-ant colony optimization algorithm for rule-based feature selection in health care

Rule-based classification in the field of health care using artificial intelligence provides solutions in decision-making problems involving different domains. An important challenge is providing access to good and fast health facilities. Cervical cancer is one of the most frequent causes of death in females. The diagnostic methods for cervical cancer used in health centers are costly and time-consuming. In this paper, bat algorithm for feature selection and ant colony optimization-based classification algorithm were applied on cervical cancer data set obtained from the repository of the University of California, Irvine to analyze the disease based on optimal features. The proposed algorithm outperforms other methods in terms of comprehensibility and obtains better results in terms of classification accuracy.

Identification of phishing websites through hyperlink analysis and rule extraction

Purpose The aim of this study is to propose an efficient rule extraction and integration approach for identifying phishing websites. The proposed approach can elucidate patterns of phishing websites and identify them accurately. Design/methodology/approach Hyperlink indicators along with URL-based features are used to build the identification model. In the proposed approach, very simple rules are first extracted based on individual features to provide meaningful and easy-to-understand rules. Then, the F-measure score is used to select high-quality rules for identifying phishing websites. To construct a reliable and promising phishing website identification model, the selected rules are integrated using a simple neural network model. Findings Experiments conducted using self-collected and benchmark data sets show that the proposed approach outperforms 16 commonly used classifiers (including seven non–rule-based and four rule-based classifiers as well as five deep learning models) in terms of interpretability and identification performance. Originality/value Investigating patterns of phishing websites based on hyperlink indicators using the efficient rule-based approach is innovative. It is not only helpful for identifying phishing websites, but also beneficial for extracting simple and understandable rules.

Smart-Cover: A real time sitting posture monitoring system

Prolonged asymmetrical sitting is common and can exacerbate musculoskeletal back pain and spinal deformities. Monitoring sitting posture can help maintain correct posture and prevent health problems. Currently posture is assessed by expert clinicians using subjective visual observation. More objective methods involve the use of a gold-standard motion capture system in Laboratories, which is expensive and not widely available. We develop a Smart-Cover, an automatic system to provide real time visualization and information about sitting posture. A Sitting Pressure Sensor (SPS) is built using Velostat, conductive fabric and foam to collect pressure distribution information within the seat surface. The data are collected from 10 healthy young subjects where each subject sits for 30 min and are transferred to a cloud server using Internet-of-Things (IoT). A rule-based classifier is used to provide timely notification to users about sitting duration and level of asymmetry. A new end-point device application is developed to show sitting balance. Dial based displays show level of asymmetry, active and static sitting, and daily summary score. Our system is user friendly, inexpensive, automatic and quantitative method for evaluating posture in real time, which has the capability of providing quantitative information about sitting behaviour. The results show that our system can be used for objective monitoring of sitting posture which has medical and social benefits. Smart-Cover can support the management of patients during their rehabilitation by monitoring pressure areas and balance when seated. In addition, it offers users at home, schools and offices the means to improve their sitting posture.

UAV Remote Sensing for Campus Monitoring: A Comparative Evaluation of Nearest Neighbor and Rule-Based Classification

UAV technology when aided with the unique data acquisition strategies, preprocessing techniques and analytical abilities of an established domain of remote sensing provide more affordable, customized and user-friendly option of “UAV-Remote Sensing”. This extended branch of remote sensing flourishes in both the mapping and measurement, if implemented in the ordered fashion to ensure remote sensing grade data. The current study integrates the potential of UAV technology to the high-resolution data classification approach of object-based image analysis. Department of Civil Engineering, Indian Institute of Technology-Roorkee, India, is selected as study area. In the first part of the study, a detailed UAV survey followed by UAV data processing was carried out to capture the VHR orthorectified image of the selected study area. In the second step, a comparative assessment of nearest neighbor (NN) and rule-based classifications were performed. Orthorectified image was segmented using a multi-resolution segmentation. The overall accuracy for NN and rule-based classifier were 95.13% and 93.87%, respectively. Detailed assessment of user accuracy and producer accuracy described that tree, road, solar panel and waterbody were more accurately classified with NN classifier, whereas building, grass land, open land and vehicle were more accurately classified with rule-based classifier.

A Multi-omic Integrative Scheme Characterizes Tissues of Action at Loci Associated with Type 2 Diabetes

SummaryResolving the molecular processes that mediate genetic risk remains a challenge because most disease-associated variants are non-coding and functional characterization of these signals requires knowledge of the specific tissues and cell-types in which they operate. To address this challenge, we developed a framework for integrating tissue-specific gene expression and epigenomic maps to obtain “tissue-of-action” (TOA) scores for each association signal by systematically partitioning posterior probabilities from Bayesian fine-mapping. We applied this scheme to credible set variants for 380 association signals from a recent GWAS meta-analysis of type 2 diabetes (T2D) in Europeans. The resulting tissue profiles underscored a predominant role for pancreatic islets and, to a lesser extent, adipose and liver, particularly among signals with greater fine-mapping resolution. We incorporated resulting TOA scores into a rule-based classifier and validated the tissue assignments through comparison with data from cis-eQTL enrichment, functional fine-mapping, RNA co-expression, and patterns of physiological association. In addition to implicating signals with a single TOA, we found evidence for signals with shared effects in multiple tissues as well as distinct tissue profiles between independent signals within heterogeneous loci. Lastly, we demonstrated that TOA scores can be directly coupled with eQTL colocalization to further resolve effector transcripts at T2D signals. This framework guides mechanistic inference by directing functional validation studies to the most relevant tissues and can gain power as fine-mapping resolution and cell-specific annotations become richer. This method is generalizable to all complex traits with relevant annotation data and is made available as an R package.

Gaussian Mixture Models for Detecting Sleep Apnea Events Using Single Oronasal Airflow Record

Sleep apnea is a common sleep-related disorder that significantly affects the population. It is characterized by repeated breathing interruption during sleep. Such events can induce hypoxia, which is a risk factor for multiple cardiovascular and cerebrovascular diseases. Polysomnography, the gold standard, is expensive, inaccessible, uncomfortable and an expert technician is needed to score sleep-related events. To address these limitations, many previous studies have proposed and implemented automatic scoring processes based on fewer sensors and machine learning classification algorithms. However, alternative device technologies developed for both home and hospital still have limited diagnostic accuracy for detecting apnea events even though many of the previous investigational algorithms are based on multiple physiological channel inputs. In this paper, we propose a new probabilistic algorithm based on (only) oronasal respiration signal for automated detection of apnea events during sleep. The proposed model leverages AASM recommendations for characterizing apnea events with respect to dynamic changes in the local respiratory airflow baseline. Unlike classical threshold-based classification models, we use a Gaussian mixture probability model for detecting sleep apnea based on the posterior probabilities of the respective events. Our results show significant improvement in the ability to detect sleep apnea events compared to a rule-based classifier that uses the same classification features and also compared to two previously published studies for automated apnea detection using the same respiratory flow signal. We use 96 sleep patients with different apnea severity levels as reflected by their Apnea-Hypopnea Index (AHI) levels. The performance was not only analyzed over obstructive sleep apnea (OSA) but also over other types of sleep apnea events including central and mixed sleep apnea (CSA, MSA). Also the performance was comprehensively analyzed and evaluated over patients with varying disease severity conditions, where it achieved an overall performance of TPR=88.5%, TNR=82.5%, and AUC=86.7%. The proposed approach contributes a new probabilistic framework for detecting sleep apnea events using a single airflow record with an improved capability to generalize over different apnea severity conditions

Identification of cut-points in test parameters for classification of industrial products

The paper considers the problem of classification of industrial products with increased quality requirements by production batches. Identification of homogeneous batches of industrial products is one of the most important problems in the way of improving the reliability and extending the lifetime. Homogeneity of product batches ensures the reliability of destructive test results. A more consistent operation of elements with identical characteristics is also provided. We solve the problem of determining the cut-point values of test parameters for further construction of a rule-based classifier.

Sarcasm Detection of Sentiments in Telugu Language

Sarcasm is usually used by people to either tease/irritate others or simply for comic purposes. The presence of sarcasm becomes certain as it is difficult to be identified by basic sentiment analysis method. Sarcasm detection is addressed with various rule-based methods, statistical approaches, and classifiers in machine learning , most of these are introduced to identify sarcasm in text written in English as it is a popular language on the internet. Although the groundwork done on sarcasm detection on various Indian languages like Telugu is limited. Hence, this paper presents a Deep learning model based on neural networks to detect sarcasm in Telugu news headlines taken from various websites . The proposed model comprises of Convolutional Neural Networks(CNN) and next a Long short-term memory(LSTM) Network which is a modified version of Recurrent neural networks (RNN) and lastly a fully connected dense layer is added to classify the sentiments into sarcastic and non-sarcastic. A pre-trained word embeddings GloVe are used in the model

Deep rule-based classifier for finger knuckle pattern recognition system

In this paper, we proposed a novel finger knuckle pattern (FKP) based personal authentication system using multilayer deep rule based (DRB) classifier. The presented approach is completely data-driven and fully automatic. However, the DRB classifier is generic and can be used in variety of classification or prediction problems. In particular, from the input finger knuckle, two kinds of features (i.e., Binarized Statistical Image Features and Gabor Filer bank) are extracted, which are then fed to fuzzy rules based DRB classifier to determine whether the user is genuine or impostor. Experimental results in the form of accuracy, error equal rate (EER) and receiver operating characteristic (ROC) curves demonstrate that presented DRB classifier is a powerful tool in FKP based biometric identification system. Experiments are reported using publicly available FKP PolyU database provided by University of Hong Kong. Experiments using this database show that the presented framework, in this study, can attain performance better than previously proposed methods. Moreover, score level fusion of all FKP modalities with BSIF + DRB yielded an equal error rate of 0.19% and an accuracy of 99.65%.

FLeAC: A Human-Centered Associative Classifier Using the Validity Concept.

Fuzzy associative classifiers (FACs) have recently received considerable attention in the data mining community due to their ability to address the imprecision and graduality of truth. Similar to their more traditional statistical peers, these classifiers, however, have remained largely data driven, not leveraging human knowledge to their advantage. This is while human expert opinion and intuition should be a unique vantage point for such systems. We introduce here, for the first time, a human-centered framework (FLeAC) for FACs based on extended fuzzy logic and f -transformation that uses experts' opinions and preferences along with statistical data to solve subjective real-world problems. In FLeAC, experts take part in both constructing and reasoning of the classifier by assigning linguistic validity to each item. These validities are then aggregated using collective intelligence that determines final item validity. To examine the proposed framework, we extend an efficient and well-known FAC, CFAR, and present an extended f -CFAR algorithm. Also, several variations of f -CFAR are implemented to examine the effect of rule validity and different f -transformation operators. We then run various nonparametric statistical tests, including Friedman, Nemenyi posthoc, and ROC tests on an actual medical dataset of burn patients from Ahwaz, Iran, to compare f -CFAR performance with those of the original and nine other rule-based classifiers. Statistical analysis shows that f -CFAR not only has a better overall diagnostic performance than CFAR but also it outperforms CFAR and the other rule-based classifiers in terms of the number of rules, the number of conditions, and the execution time, leading to a more compact and comprehensible classifier with comparable accuracy.

A Technique to Calculate National Happiness Index by Analyzing Roman Urdu Messages Posted on Social Media

National Happiness Index (NHI) is a national indicator of development that estimates the economic and social well-being of the nation's individuals. With the proliferation of the internet, people share a significant amount of data on social media websites. We can process the data with different sentiment analysis techniques to calculate the NHI. In the literature, different approaches have been used to calculate NHI, which include the lexicon-based approach and machine learning approach. All of these existing approaches are proposed to calculate NHI for the sentiments written in the English language. However, these methods fail for complex Roman Urdu tweets that contain more than two sub-opinions. There are three primary objectives of the research: (1) to investigate current sentiment analysis techniques are sufficient for the classification of complex Roman Urdu sentiments; (2) to propose rule-based classifier for the classification of Roman Urdu sentiments comprising multiple sub-opinions; (3) to calculate NHI using Roman Urdu sentiments. For this purpose, we proposed the discourse information extractor, the rule-based method (3-RBC), and the machine learning classifier. The experimental results show that 3-RBC is efficient for feature identification, and it is more statistically significant than the baseline classifiers. The 3-RBC has successfully increased the accuracy by 7% and precision by 8%, which provides evidence that the proposed technique significantly increased the calculation of NHI.

A Comparative Performance Evaluation of Classification Algorithms for Clinical Decision Support Systems

Classification algorithms are widely taken into account for clinical decision support systems. However, it is not always straightforward to understand the behavior of such algorithms on a multiple disease prediction task. When a new classifier is introduced, we, in most cases, will ask ourselves whether the classifier performs well on a particular clinical dataset or not. The decision to utilize classifiers mostly relies upon the type of data and classification task, thus making it often made arbitrarily. In this study, a comparative evaluation of a wide-array classifier pertaining to six different families, i.e., tree, ensemble, neural, probability, discriminant, and rule-based classifiers are dealt with. A number of real-world publicly datasets ranging from different diseases are taken into account in the experiment in order to demonstrate the generalizability of the classifiers in multiple disease prediction. A total of 25 classifiers, 14 datasets, and three different resampling techniques are explored. This study reveals that the classifier that is likely to become the best performer is the conditional inference tree forest (cforest), followed by linear discriminant analysis, generalize linear model, random forest, and Gaussian process classifier. This work contributes to existing literature regarding a thorough benchmark of classification algorithms for multiple diseases prediction.

Artificial intelligence applied to small businesses: the use of automatic feature engineering and machine learning for more accurate planning

The purpose of this study is to develop a predictive model that increases the accuracy of business operational planning using data from a small business. By using Machine Learning (ML) techniques feature expansion, resampling, and combination techniques, it was possible to address several existing limitations in the available research. Then, the use of the novel technique of feature engineering allowed us to increase the accuracy of the model by finding 10 new features derived from the original ones and constructed automatically through the nonlinear relationships found between them. Finally, we built a rule-based classifier to predict the store's revenue with high accuracy. The results show the proposed approach open new possibilities for ML research applied to small and medium businesses.

Applying a rule-based object-based image analysis approach for nearshore bar identification and characterization

Nearshore bars are important coastal features that influence beach–dune dynamics and protect the coast from high-energy wave events. Historically, in situ methods were used to study beach–dune dynamics and nearshore bars. Video monitoring systems, such as Argus, have been the favored remote sensing approach for studying nearshore bars dynamics and evolution. These systems are costly to install and maintain, and a limited number of stations are available. Our study introduces a lower cost alternative to study these coastal features. A rule-based classification object-based image analysis approach is introduced to identify nearshore bars (and their characteristics, such as bar length, width, count, and distance from the wet/dry line) from very-high-resolution (VHR) multispectral imagery. Nearshore bars were successfully identified and characterized as single and multiple bar systems. Our study demonstrates that VHR satellite remote sensing is a viable alternative with comparative results to video monitoring systems. Their capability of acquiring high-resolution imagery in large spatial extent is superior in continuous observation and characterization of nearshore bars. The proposed approach provides information at a more relevant spatiotemporal scale at which these systems evolve, which is helpful to coastal scientists and managers.