Similarity-based Classification Research Articles

Automated Identification of Patients’ Unmet Social Needs in Clinical Text Using Natural Language Processing

ObjectiveTo develop natural language processing (NLP) solutions for identifying patients' unmet social needs to enable timely intervention.Patients and Methods: DesignRetrospective cohort study with review and annotation of clinical notes to identify unmet social needs, followed by using the annotations to develop and evaluate NLP solutions. Participants1,103 primary care patients seen at a large academic medical center during 6/1/2019-5/31/2021 and referred to a community health worker (CHW program). Clinical notes and portal messages of 200 age/sex-stratified patients were sampled for annotation of unmet social needs. SystemsTwo NLP solutions were developed and compared. The first solution employed similarity-based classification on top of sentences represented as semantic embedding vectors. The second solution involved designing of terms and patterns for identifying each domain of unmet social needs in the clinical text. MeasuresPrecision, recall, and f1-score of the NLP solutions. ResultsA total of 5,675 clinical notes and 475 portal messages were annotated, with an inter-annotator agreement of 0.938. The best NLP solution achieved an f1-score of 0.95 and was applied to the entire CHW-referred cohort (n=1,103), of whom >80% had at least one unmet social need within the 6 months prior to the first CHW referral. Financial strain and health literacy were the top two domains of unmet social needs across most of the sex/age strata. ConclusionClinical text contains rich information about patients’ unmet social needs. NLP can achieve good performance in identifying those needs for CHW referral and facilitate data-driven research on social determinants of health.

A Novel Approach for Mining Big Data Using Multi-Model Fusion Mechanism (MMFM)

Big data processing and analytics require sophisticated systems and cutting-edge methodologies to extract useful data from the available data. Extracted data visualization is challenging because of the processing models' dependence on semantics and classification. To categorize and improve information-based semantics that have accumulated over time, this paper introduces the Multi-model fusion mechanism for data mining (MMFM) approach. Information dependencies are organized based on the links between the data model based on attribute values. This method divides the attributes under consideration based on processing time to handle complicated data in controlled amount of time. The proposed MMFM’s performance is assessed with real-time weather prediction dataset where the data is acquired from sensor (observed) and image data. MMFM is used to conduct semantic analytics and similarity-based classification on this collection. The processing time based on records and samples are investigated for the various data sizes, instances, and entries. It is found that the proposed MMFM gets 70 seconds of processing time for 2GB data and 0.99 seconds while handling 5000 records for various classification instances.

MODULATION FOR INFORMATION MEASUREMENT SYSTEM IN CONTEXT OF SOFTWARE DEVELOPMENT USING NODE-BASED GRAPHICAL EDITOR

In this paper information measurement system (IMS) modulation integrated with a graphical node editor (GNE) is carried out for engineering applications to build software capable of both editing and simulating different scenarios. A modulation characteristic measuring and classification system for classifying a discreet signal that includes a preprocessor, stage measuring system and modulation classifier. Preprocessor detects impulse begins and finishes, determines impulse time, and changes the discreet signal into digital signals. The stage measuring system determines short chip counts, long chip counts, stage jump magnitudes, number of stage states, and polynomial coefficients for stage modulation of the discreet signal. Modulation classifier identifies modulation type on base of measurements using both rules-based and similarity-based classification methods. The lack of educational-oriented software in the IMS field leads to a more challenging learning curve resulting in a drop in student motivation. Software integrated with GNE is purposed to minimize the overwhelming feeling in students in the IMS field by simultaneously providing understandable execution order for the machine. An application consisting of GNE with the addition of custom node architecture and their interaction logic is built to conduct laboratory work using this software. Executed laboratory work result shows that having GNE integrated into the software can significantly impact visual understanding, leading to higher motivation and interest in the subject in students. Keywords: Node, Graphical Editor, GPIOS, Information measurement system.

Five themes for ecostylistics

Abstract Fundamentally, meaning is organized along two dimensions, similarity and contiguity, corresponding to two areas of the brain primarily responsible for language processing, Wernicke’s area and Broca’s area, respectively. Modern culture has tended to overemphasize the similarity dimension through money/commodification in capitalism, and mathematics in science, with disastrous ecological consequences. Ecostylistics can celebrate and analyze themes and linguistic patterns of poetry and novels which challenge this overemphasis. Five such themes are suggested in this article. To counter overemphasis on similarity (1) individuation. To celebrate Broca’s area’s contiguity dimension (2) dynamic process, and (3) interrelatedness and communication with the natural world. However, concentration on local contiguities of time, manifest in the contemporary English-speaking novel, distracts from the global contiguities of (4) long-term ecological change. (5) The two dimensions are also manifest in metaphor, which challenges conventional similarity-based classification, and narrative, which expands the contiguity dimension beyond the clause. These themes are illustrated by poetic examples from Wordsworth and Edward Thomas, and detailed analysis of the following texts: Gerard Manley Hopkins’ “As Kingfishers Catch Fire” and “Nature is a Heraclitean Fire”; Edward Thomas’s “The Word”, Alice Oswald’s “Sonnet” and “Birdsong for Two Voices”, and, returning to a text whose analysis by Michael Halliday was seminal for Stylistics, William Golding’s The Inheritors. Transitivity analysis within the framework of Systemic Functional Grammar is used throughout, and connections are made with quantum physics, Daoism, and other process philosophies.

Synthetic Lethal Interactions Prediction Based on Multiple Similarity Measures Fusion

The synthetic lethality (SL) relationship arises when a combination of deficiencies in two genes leads to cell death, whereas a deficiency in either one of the two genes does not. The survival of the mutant tumor cells depends on the SL partners of the mutant gene, thereby the cancer cells could be selectively killed by inhibiting the SL partners of the oncogenic genes but normal cells could not. Therefore, there is an urgent need to develop more efficient computational methods of SL pairs identification for cancer targeted therapy. In this paper, we propose a new approach based on similarity fusion to predict SL pairs. Multiple types of gene similarity measures are integrated and k-nearest neighbors algorithm (k-NN) is applied to achieve the similarity-based classification task between gene pairs. As a similarity-based method, our method demonstrated excellent performance in multiple experiments. Besides the effectiveness of our method, the ease of use and expansibility can also make our method more widely used in practice.

Few-Shot Open World Learner

Abstract Computer vision based recognition systems in dynamically changing environments require continuously updating datasets with novel detected categories while maintaining equally high performance on previously established classes. These requirements can be addressed by the concept of Open World Learning introduced by (Chen et al., 2018). We propose a novel framework as a solution to the open world learning problem. This solution is based on the few-shot classification strategy in combination with an outlier detection module. The few-shot classifiers utilized in this work provide a similarity-based classification scheme and present an excellent solution to the incremental learning problem natively, requiring no training and only a few labeled examples of new classes before adapting to them. The discovery of new classes is performed by an outlier detection module that utilizes the similarity space created by the few-shot classifier to identify sufficiently different samples from the known classes and remove them from the classification process. This paper highlights the best combination of few-shot and outlier detection algorithms. Extensive experiments with different combinations of these algorithms are conducted in the pursuit of the ideal components of the framework. Results of the experiments show that we are successful in creating a novel implementation of an open world learner with a very limited loss of accuracy compared to the base few-shot algorithm.

Comparing tagging suggestion models on discrete corpora

Purpose This paper aims to investigate the methods for the prediction of tags on a textual corpus that describes diverse data sets based on short messages; as an example, the authors demonstrate the usage of methods based on hotel staff inputs in a ticketing system as well as the publicly available StackOverflow corpus. The aim is to improve the tagging process and find the most suitable method for suggesting tags for a new text entry. Design/methodology/approach The paper consists of two parts: exploration of existing sample data, which includes statistical analysis and visualisation of the data to provide an overview, and evaluation of tag prediction approaches. The authors have included different approaches from different research fields to cover a broad spectrum of possible solutions. As a result, the authors have tested a machine learning model for multi-label classification (using gradient boosting), a statistical approach (using frequency heuristics) and three similarity-based classification approaches (nearest centroid, k-nearest neighbours (k-NN) and naive Bayes). The experiment that compares the approaches uses recall to measure the quality of results. Finally, the authors provide a recommendation of the modelling approach that produces the best accuracy in terms of tag prediction on the sample data. Findings The authors have calculated the performance of each method against the test data set by measuring recall. The authors show recall for each method with different features (except for frequency heuristics, which does not provide the option to add additional features) for the dmbook pro and StackOverflow data sets. k-NN clearly provides the best recall. As k-NN turned out to provide the best results, the authors have performed further experiments with values of k from 1–10. This helped us to observe the impact of the number of neighbours used on the performance and to identify the best value for k. Originality/value The value and originality of the paper are given by extensive experiments with several methods from different domains. The authors have used probabilistic methods, such as naive Bayes, statistical methods, such as frequency heuristics, and similarity approaches, such as k-NN. Furthermore, the authors have produced results on an industrial-scale data set that has been provided by a company and used directly in their project, as well as a community-based data set with a large amount of data and dimensionality. The study results can be used to select a model based on diverse corpora for a specific use case, taking into account advantages and disadvantages when applying the model to your data.

AEAO: Auto Encoder with Adam Optimizer Method for Efficient Document Indexing of Big Data

In the big data era, the document classification became an active research area due to the explosive nature in the volumes of data. Document Indexing is one of the important tasks under text classification. The objective of this research is to increase the performance of the document indexing by proposing Adam optimizer in the auto-encoder. Due to the larger dimension and multi-class classification problem, the accuracy of document indexing is reduced. In this paper, an enhanced auto encoder is used based on the objective function of the Adam optimization (AEAO), which improves the learning rate and accuracy of indexing. The documents from the 20-newsgroup data set are converted into vector representation, and then the cosine similarity and Pearson correlation have been measured from the vector. The word to vector representation has words in the vector form and the frequency of words in the document increases their value. The Adam optimization technique selects the features by using similarity values and improves the learning rate. The auto encoder classifier classifies the document based on the objective function of the Adam optimizer. The experiment is conducted using python and the result infers that the classification performance of AEAO is better than that of Similarity-based classification framework for Multiple-Instance Learning and Self-Adaptive LSH encoding for multi-instance Learning techniques in terms of parameters like precision, recall and f-score.

A Robust Distance Measure for Similarity-Based Classification on the SPD Manifold.

The symmetric positive definite (SPD) matrices, forming a Riemannian manifold, are commonly used as visual representations. The non-Euclidean geometry of the manifold often makes developing learning algorithms (e.g., classifiers) difficult and complicated. The concept of similarity-based learning has been shown to be effective to address various problems on SPD manifolds. This is mainly because the similarity-based algorithms are agnostic to the geometry and purely work based on the notion of similarities/distances. However, existing similarity-based models on SPD manifolds opt for holistic representations, ignoring characteristics of information captured by SPD matrices. To circumvent this limitation, we propose a novel SPD distance measure for the similarity-based algorithm. Specifically, we introduce the concept of point-to-set transformation, which enables us to learn multiple lower dimensional and discriminative SPD manifolds from a higher dimensional one. For lower dimensional SPD manifolds obtained by the point-to-set transformation, we propose a tailored set-to-set distance measure by making use of the family of alpha-beta divergences. We further propose to learn the point-to-set transformation and the set-to-set distance measure jointly, yielding a powerful similarity-based algorithm on SPD manifolds. Our thorough evaluations on several visual recognition tasks (e.g., action classification and face recognition) suggest that our algorithm comfortably outperforms various state-of-the-art algorithms.

Genotyping genetically heterogeneous Cyclospora cayetanensis infections to complement epidemiological case linkage.

Sexually reproducing pathogens such as Cyclospora cayetanensis often produce genetically heterogeneous infections where the number of unique sequence types detected at any given locus varies depending on which locus is sequenced. The genotypes assigned to these infections quickly become complex when additional loci are analysed. This genetic heterogeneity confounds the utility of traditional sequence-typing and phylogenetic approaches for aiding epidemiological trace-back, and requires new methods to address this complexity. Here, we describe an ensemble of two similarity-based classification algorithms, including a Bayesian and heuristic component that infer the relatedness of C. cayetanensis infections. The ensemble requires a set of haplotypes as input and assigns arbitrary distances to specimen pairs reflecting their most likely relationships. The approach was applied to data generated from a test cohort of 88 human fecal specimens containing C. cayetanensis, including 30 from patients whose infections were associated with epidemiologically defined outbreak clusters of cyclosporiasis. The ensemble assigned specimens to plausible clusters of genetically related infections despite their complex haplotype composition. These relationships were corroborated by a significant number of epidemiological linkages (P < 0.0001) suggesting the ensemble's utility for aiding epidemiological trace-back investigations of cyclosporiasis.

Adaptive Model-Based Classification of PolSAR Data

An adaptive classification is developed as a hybrid of the eigenvector-based and the model-based target decompositions for polarimetric synthetic aperture radar (PolSAR) data. The classification adopts the canonical scattering models that widely used in model-based decompositions to provide an improvement for the well-known ${H/\alpha }$ classification. First, a correspondence principle is adopted to adaptively identify the matched canonical models. The selected models are parallelly combined based on the scattering similarity for a fine depiction of the scattering mechanism then. Twelve classes are finally obtained, and each one carries a unique symbol to show a specific scattering. The classification does not depend on a particular data set, avoids the hard partitioning, and solves the obscures in ${H/\alpha }$ . Comparison on the real PolSAR data sets with ${H/\alpha }$ and the existing scattering similarity-based classification validates the better discrimination.

LVQ-KNN: Composition-based DNA/RNA binning of short nucleotide sequences utilizing a prototype-based k-nearest neighbor approach

Unbiased sequencing is an upcoming method to gain information of the microbiome in a sample and for the detection of unrecognized pathogens. There are many software tools for a taxonomic classification of such metagenomics datasets available. Numerous of them have a satisfactory sensitivity and specificity for known organisms, but they fail if the sample contains unknown organisms, which cannot be detected by similarity-based classification employing available databases. However, recognition of unknowns is especially important for the detection of newly emerging pathogens, which are often RNA viruses. Here we present the composition-based analysis tool LVQ-KNN for binning unclassified nucleotide sequence reads into their provenance classes DNA or RNA. With a 5-fold cross-validation, LVQ-KNN reached correct classification rates (CCR) of up to 99.9% for the classification into DNA/RNA. Real datasets gained CCRs of up to 94.5%. Comparing the method to another composition-based analysis tool, similar or better classification results were reached. LVQ-KNN is a new tool for DNA/RNA classification of sequence reads from unbiased sequencing approaches that could be applicable for the detection of yet unknown RNA viruses in metagenomic samples. The source-code, training and test data for LVQ-KNN is available at Github (https://github.com/ab1989/LVQ-KNN).

Algorithmic sign prediction and covariate selection across eleven international stock markets

I investigate whether an expert system can be used for profitable long-term asset management. The trading strategy of the expert system needs to be based on market predictions. To this end, I generate binary predictions of the market returns by using statistical and machine-learning algorithms. The methods used include logistic regressions, regularized logistic regressions and similarity-based classification. I test the methods in a contemporary data set involving data from eleven developed markets. Both statistical and economic significance of the results are considered. As an ensemble, the results seem to indicate that there is some degree of mild predictability in the stock markets. Some of the results obtained are highly significant in the economic sense, featuring annualized excess returns of 3.1% (France), 2.9% (Netherlands) and 0.8% (United States). However, statistically significant results are seldom found. Consequently, the results do not completely invalidate the efficient-market hypothesis.

Predicting drug-disease associations and their therapeutic function based on the drug-disease association bipartite network.

Drug-disease associations provide important information for drug discovery and drug repositioning. Drug-disease associations can induce different effects, and the therapeutic effect attracts wide spread interest. Therefore, developing drug-disease association prediction methods is an important task, and differentiating therapeutic associations from other associations is also very important. In this paper, we formulate the known drug-disease associations as a bipartite network, and then present a novel representation for drugs and diseases based on the bipartite network and linear neighborhood similarity. Thus, we propose the network topological similarity-based inference method (NTSIM) to predict unobserved drug-disease associations. Further, we extend the work to the association classification, and propose the network topological similarity-based classification method (NTSIM-C) to differentiate therapeutic associations from others. Compared with existing drug-disease association prediction methods, NTSIM can produce superior performances in predicting drug-disease associations, and NTSIM-C can accurately classify drug-disease associations. Further, we analyze the capability of proposed methods by using several case studies. The studies show the usefulness of NTSIM and NTSIM-C in the real applications. In conclusion, NTSIM and NTSIM-C are promising for predicting drug-disease associations and their therapeutic functions.

Poisonous Shrimp Detection System for Litopenaeus Vannamei using k-Nearest Neighbor Method

One of the important seafoods in the food consumption of humans is shrimp. Although shrimp contains proteins that are needed by the human body, sometimes it contains toxins. This is due to environmental factors or catching processes that may use toxins. Therefore, the community should take precautions when consuming shrimp. White shrimp (Litopenaeus vannamei) is one type of shrimp that is preferred because of its delicious taste. The purpose of this research is to develop a computerized system for poisonous white shrimp detection. The category of white shrimps consists of two kinds, i.e., fresh white shrimps that are caught in a natural way (class A), and poisonous white shrimps that are caught by using toxin (class B). The features used are RGB colors (red, green, and blue) and texture (energy, contrast, correlation, and homogeneity). A similarity-based classification is performed by the k-Nearest Neighbor (k-NN) algorithm. The experiment was conducted on a dataset consisting of 90 white shrimp images. The holdout validation method was used to evaluate the system. The original dataset was divided into two parts, whereby 60 images were used as training samples and 30 images were used as testing images. Based on the evaluation results, it can be concluded that the classification accuracy is 73.33%. The benefit of the developed system is to help the community in selecting good and safe white shrimps.

Fuzzy similarity-based classification method for gender recognition using 3D facial images

In this paper, we propose a new fuzzy similarity-based classification (FSBC) method for the task of gender recognition. The proposed method characterises each individual by extracting geometrical features from a 3D facial image using pertinent radial curves. Our approach includes representing the extracted features using fuzzy sets to handle imprecision in its values. Also, the proposed FSBC method recognises the gender of a new person by evaluating his similarity to the male and female samples pre-set as gender representatives set, then we aggregate the obtained similarities to compute the scores of belonging to each gender. In the end, we ascribe to each new person the gender with the higher score. With the proposed method, two main advantages are obtained: First, we used the OWA operator and RIM quantifier to define the percentage of significant features for the similarity assessment. Second, the aggregation process was performed using compensatory operators to ensure the selected gender has high similarities. Experiments were conducted using FRAV3D database, by considering only one frontal pose in the gender representatives set. The obtained gender recognition rate of the proposed method was very promising compared to other classification method.

Fuzzy similarity-based classification method for gender recognition using 3D facial images

In this paper, we propose a new fuzzy similarity-based classification (FSBC) method for the task of gender recognition. The proposed method characterises each individual by extracting geometrical features from a 3D facial image using pertinent radial curves. Our approach includes representing the extracted features using fuzzy sets to handle imprecision in its values. Also, the proposed FSBC method recognises the gender of a new person by evaluating his similarity to the male and female samples pre-set as gender representatives set, then we aggregate the obtained similarities to compute the scores of belonging to each gender. In the end, we ascribe to each new person the gender with the higher score. With the proposed method, two main advantages are obtained: First, we used the OWA operator and RIM quantifier to define the percentage of significant features for the similarity assessment. Second, the aggregation process was performed using compensatory operators to ensure the selected gender has high similarities. Experiments were conducted using FRAV3D database, by considering only one frontal pose in the gender representatives set. The obtained gender recognition rate of the proposed method was very promising compared to other classification method.

Malware classification using dynamic features and Hidden Markov Model

In recent years the number of new malware threats has increased significantly, causing a damage of billions of dollars globally. To counter this aggressive malware attack, the anti-malware industry needs to be able to correctly classify malware in order to provide defense against them. Consequently, malware classification has been an active area of research, and a multitude of malware classification approaches have been proposed in the literature. This paper evaluates two methods of sequence classification based on Hidden Markov Model, namely the maximum likelihood and similarity-based methods, for classification of malware using a large and comprehensive dataset. System calls generated by known malware during execution are used as observation sequences to train the Hidden Markov Models. Malware samples are evaluated against the trained models to produce similarity vectors, which are used in the maximum likelihood and similarity-based classification schemes to predict the family for an unknown malware sample. Comparison of the two schemes shows that combining the powerful statistical pattern analysis capability of Hidden Markov Models and discriminative classifiers in the similarity- based method results in a significantly better classification performance as compared to the maximum likelihood approach. Furthermore, evaluation of different classifiers in the similarity-based method demonstrates that Random Forest classifier performs better than other classifiers on malware similarity vectors.

Exploring similarity-based classification of larynx disorders from human voice

In this paper identification of laryngeal disorders using cepstral parameters of human voice is researched. Mel-frequency cepstral coefficients (MFCCs), extracted from audio recordings of patient’s voice, are further approximated, using various strategies (sampling, averaging, and clustering by Gaussian mixture model). The effectiveness of similarity-based classification techniques in categorizing such pre-processed data into normal voice, nodular, and diffuse vocal fold lesion classes is explored and schemes to combine binary decisions of support vector machines (SVMs) are evaluated. Most practiced RBF kernel was compared to several constructed custom kernels: (i) a sequence kernel, defined over a pair of matrices, rather than over a pair of vectors and calculating the kernelized principal angle (KPA) between subspaces; (ii) a simple supervector kernel using only means of patient’s GMM; (iii) two distance kernels, specifically tailored to exploit covariance matrices of GMM and using the approximation of the Kullback–Leibler divergence from the Monte-Carlo sampling (KL-MCS), and the Kullback–Leibler divergence combined with the Earth mover’s distance (KL-EMD) as similarity metrics.The sequence kernel and the distance kernels both outperformed the popular RBF kernel, but the difference is statistically significant only in the distance kernels case. When tested on voice recordings, collected from 410 subjects (130 normal voice, 140 diffuse, and 140 nodular vocal fold lesions), the KL-MCS kernel, using GMM with full covariance matrices, and the KL-EMD kernel, using GMM with diagonal covariance matrices, provided the best overall performance. In most cases, SVM reached higher accuracy than least squares SVM, except for common binary classification using distance kernels. The results indicate that features, modeled with GMM, and kernel methods, exploiting this information, is an interesting fusion of generative (probabilistic) and discriminative (hyperplane) models for similarity-based classification.

Similarity-Based Classification in Relational Databases

In this paper, we introduce a method for measuring similarity of objects of a relational database (relational objects, in short). We also propose and investigate an algorithm SC for classification of relational objects. The task of classification is carried out based on similarity of the objects to predefined classes. An object to be classified is assigned to the class to which it is most similar. A similarity of an object to a class is understood as its similarity to a class representative. Several methods for computing the class representative are proposed. We test the algorithm on real and artificial databases. We compare results obtained by the algorithm with those obtained by other algorithms known from the literature. We also present our approach in the context of granular computing.