K-nearest Neighbor Classification Technique Research Articles

Predicting Mental Disorders Using Data Mining Techniques

Over the last few years, data mining makes a transition to new applications in medicine. Many people suffer from mental disorders because of absence of care in the health care sector, especially in the mental care. Lack of health records for patients of mental health, has been caused by many reasons, for example, patients’ hesitation to go the psychiatrist to follow-up their condition due high cost of examination fee, waiting time for their turn, and diagnosis inaccuracy. Classifiers are used to predict normal, addiction without disorder, addiction with disorder, or disorder without addiction patients’ cases. However, a rise in the cases of personality disorder and substance abuse. We used Naive Bayes and K-Nearest Neighbor classification technique on dataset consists of opioid addicts patients derived from the United States of America to obtain high accuracy of mental disorder prediction. The precision of the NB classifier is (87%), and the precision of KNN classifier is (92%). This paper presents a new method for diagnosing mental disorders through data mining.

Detection and Classification of Caterpillar using Image Processing with K-Nearest Neighbor Classification Technique

Caterpillars are the developmental stage of the flying insect called butterfly. The moths are the beautiful creature of earth which comes under the class of insects. They are recognized by their beautiful and colorful forewings body and legs. Caterpillars are the larval stage of the moth which are found in the leaf and stem of the plants when the moth laid eggs on the leaves after their mating. Caterpillar after fully developed from its eggs draw out a flimsy, soft cocoon made up of dark coarse silk on leaves and stem for their shelter. Caterpillar is also a beautiful creature that is found with different colors and strips with spines and urticating hair in their body for releasing venom for self-defense from external predators. The present study works on the detection and classification of the caterpillar using image processing with a k-NN classifier.This research help in characterizing the type of caterpillar image classification for particular three classes such as accuracy of Buck Moth Caterpillar, the accuracy of Saddleback Caterpillar, and the accuracy of Io moth Caterpillar. The following stages have been considered are preprocessing, segmentation, feature extraction, and classification methods using K- Nearest Neighbor classifier. The present investigation results that SYMLET5 analysis works well in the classification of the caterpillar with an accuracy of 96% using K- Nearest Neighbor classifier compare with other measures during investigation and analysis.

SLOPE AND CURVE CODING TO RECOGNIZE ENGLISH NUMERALS

Purpose: Lines and Curves are important parts of characters in any script. Features based on lines and curves go a long way to characterize an individual character as well as differentiate similar-looking characters. The present paper proposes an English numerals recognition system using feature elements obtained from the novel and efficient coding of the curves and local slopes. The purpose of this paper is to recognize English numerals efficiently to develop a reliable Optical Character recognition system.
 Methodology: K-Nearest Neighbour classification technique has been implemented on a global database MNIST to get an overall recognition accuracy rate of 96.7 %, which is competitive to other reported works in literature. Distance features and slope features are extracted from pre-processed images. The feature elements from training images are used to train K-Nearest-Neighbour classifier and those from test images have been used to classify them.
 Main Findings: The findings of the current paper can be used in Optical Character Recognition (OCR) of alphanumeric characters of any language, automatic reading of amount on bank cheque, address written on envelops, etc.
 Implications: Due to the similarity in structures of some numerals like 2, 3, and 8, the system produces respectively lower recognition accuracy rates for them.
 Novelty: The ways of finding distance and slope features to differentiate the curves in the structure of English Numerals is the novelty of this work.

ARSkNN: An efficient k-nearest neighbor classification technique using mass based similarity measure

ARSkNN: An efficient k-nearest neighbor classification technique using mass based similarity measure

SELEÇÃO DE VARIÁVEIS PARA CATEGORIZAÇÃO DE AMOSTRAS QUÍMICAS

This paper presents a method to select the best variables to categorize chemical samples into two classes, say conforming or non-conforming. For that matter, PLS regression is combined with a data mining tool, the k-Nearest Neighbor classification technique, through an iterative variable selection process. The recommended subset of variables is chosen based on several criteria: sensitivity, specificity and percent of retained variables. When applied to two datasets related to wine analysis and one associated to QSAR, the proposed method significantly reduced the number of variables required for classification, while yielding superior categorization performance when compared to using all original variables.

Variable selection framework for allocating products to recommended replenishment models in VMI applications

PurposeThe purpose of this paper is to propose a multivariate-based method to classify products in replenishment categories based on principal component analysis (PCA) along with two classification algorithms, k-nearest neighbor (KNN) and linear discriminant analysis (LDA).Design/methodology/approachIn the propositions, PCA is applied to data describing products’ features and demand behavior, and a variable importance index (VII) is derived based on PCA parameters. Next, products are allocated to inventory replenishment models applying the KNNs to all original variables; the classification accuracy is then assessed. The variable with the smallest VII is removed and a new classification is carried out; this iterative procedure is performed until a single variable is left. The subset yielding the maximum classification accuracy is recommended for future classification. The aforementioned procedure is repeated replacing the KNN by the LDA.FindingsWhen applied to real data from a consulting company, the KNN classification technique led to higher performance levels than LDA, yielding 89.4 percent average accuracy and retaining about 80 percent of the original variables. On the other hand, LDA reached 87.1 percent average accuracy and retained 95 percent of the variables. Based on such results, the authors’ findings suggest that 14 out of the 24 variables are crucial in determining an inventory replenishment model for a product in a specific location replacement. Several of the retained variables were typically used in reorder point estimation or associated to market profile in specific locals.Originality/valueThe idea of this paper is to remove irrelevant and noisy market metrics that jeopardize the correct allocation of products to the most appropriate replenishment model.

A genetic algorithm-based framework for wavelength selection on sample categorization.

In forensic and pharmaceutical scenarios, the application of chemometrics and optimization techniques has unveiled common and peculiar features of seized medicine and drug samples, helping investigative forces to track illegal operations. This paper proposes a novel framework aimed at identifying relevant subsets of attenuated total reflectance Fourier transform infrared (ATR-FTIR) wavelengths for classifying samples into two classes, for example authentic or forged categories in case of medicines, or salt or base form in cocaine analysis. In the first step of the framework, the ATR-FTIR spectra were partitioned into equidistant intervals and the k-nearest neighbour (KNN) classification technique was applied to each interval to insert samples into proper classes. In the next step, selected intervals were refined through the genetic algorithm (GA) by identifying a limited number of wavelengths from the intervals previously selected aimed at maximizing classification accuracy. When applied to Cialis®, Viagra®, and cocaine ATR-FTIR datasets, the proposed method substantially decreased the number of wavelengths needed to categorize, and increased the classification accuracy. From a practical perspective, the proposed method provides investigative forces with valuable information towards monitoring illegal production of drugs and medicines. In addition, focusing on a reduced subset of wavelengths allows the development of portable devices capable of testing the authenticity of samples during police checking events, avoiding the need for later laboratorial analyses and reducing equipment expenses. Theoretically, the proposed GA-based approach yields more refined solutions than the current methods relying on interval approaches, which tend to insert irrelevant wavelengths in the retained intervals. Copyright © 2016 John Wiley & Sons, Ltd.

Detecting homogenous clusters using whole-rock chemical compositions and REE patterns: A graph-based geochemical approach

The rock chemical composition may be affected by a wide variety of primary and secondary geological processes. Analyzing geochemical datasets of highly altered rocks is usually faced with the challenges in detecting the relationships among objects. Using traditional clustering methods in such datasets with high-dimensional data and various types of attributes commonly leads to poor quality results. Hence, a graph-based geochemical approach was proposed in this study to solve this problem. In order to determine the relationship between objects, various similarity measures related to whole-rock composition, REE pattern, and the geographical location were employed in combination to weight the edges of a similarity graph. A spectral method was effectively used to identify clusters (communities) representing rock groups, geological units, or geochemical zones in the weighted similarity graph. It could also recognize the corresponding groups being compositionally and genetically similar to each other and distinguish sub-groups or anomalous samples in the dataset with regard to the different levels of clustering. Firstly, the performance and effectiveness of the proposed approach was evaluated by testing on a GEOROC11Geochemistry of Rocks of the Oceans and Continents. dataset based on some graph clustering quality functions. Then the approach was applied to the geochemical dataset of Choghart orebody comprising various altered rocks. The obtained clusters were visualized by a k-nearest neighbor classification technique to represent geochemical zones as a continuous map.

ARSkNN-A k-NN Classifier Using Mass Based Similarity Measure

k-Nearest Neighbor (k-NN) classification technique is one of the most elementary and straightforward classification methods. Although distance learning is in the core of k-NN classification, similarity can be preferred upon distance in several practical applications. This paper proposes a novel algorithm for learning a class of an instance based on a similarity measure which does not calculate distance, for k-Nearest Neighbor (k-NN) classification.

Machine Learning Techniques for Cooperative Spectrum Sensing in Cognitive Radio Networks

We propose novel cooperative spectrum sensing (CSS) algorithms for cognitive radio (CR) networks based on machine learning techniques which are used for pattern classification. In this regard, unsupervised (e.g., K-means clustering and Gaussian mixture model (GMM)) and supervised (e.g., support vector machine (SVM) and weighted K-nearest-neighbor (KNN)) learning-based classification techniques are implemented for CSS. For a radio channel, the vector of the energy levels estimated at CR devices is treated as a feature vector and fed into a classifier to decide whether the is available or not. The classifier categorizes each feature vector into either of the two classes, namely, the channel available and the channel unavailable class. Prior to the online classification, the classifier needs to go through a training phase. For classification, the K-means clustering algorithm partitions the training feature vectors into K clusters, where each cluster corresponds to a combined state of primary users (PUs) and then the classifier determines the class the test energy vector belongs to. The GMM obtains a mixture of Gaussian density functions that well describes the training feature vectors. In the case of the SVM, the support vectors (i.e., a subset of training vectors which fully specify the decision function) are obtained by maximizing the margin between the separating hyperplane and the training feature vectors. Furthermore, the weighted KNN classification technique is proposed for CSS for which the weight of each feature vector is calculated by evaluating the area under the receiver operating characteristic (ROC) curve of that feature vector. The performance of each classification technique is quantified in terms of the average training time, the sample classification delay, and the ROC curve. Our comparative results clearly reveal that the proposed algorithms outperform the existing state-of-the-art CSS techniques.

An Alternative Algorithm for Classification Based on Robust Mahalanobis Distance

This study considers the classification problem for binary output attribute when input attributes are drawn from multivariate normal distribution, in both clean and contaminated case. Classical metrics are affected by the outliers, while robust metrics are computationally inefficient. In order to achieve robustness and computational efficiency at the same time, we propose a new robust distance metric for K-Nearest Neighbor (KNN) method. We call our proposed metric Alternative Robust Mahalanobis Distance (ARMD) metric. Thus KNN using ARMD is alternative KNN method. The classical metrics use non robust estimate (mean) as the building block. To construct the proposed ARMD metric, we replace non robust estimate (mean) by its robust counterpart median. Thus, we developed ARMD metric for alternative KNN classification technique. Our simulation studies show that the proposed alternative KNN method gives better results in case of contaminated data compared to the classical KNN. The performance of our method is similar to classical KNN using the existing robust metric. The major advantage of proposed method is that it requires less computing time compared to classical KNN that using existing robust metric. Dhaka Univ. J. Sci. 61(1): 81-85, 2013 (January) DOI: http://dx.doi.org/10.3329/dujs.v61i1.15101

A multivariate-based wavenumber selection method for classifying medicines into authentic or counterfeit classes

Attenuated total reflectance (ATR), a sampling technique by Fourier transform infrared (FTIR) spectroscopy, has been adopted as an analytical tool for detecting fraudulent medicines. The spectrum generated by FTIR-ATR typically relies on hundreds of equally spaced wavenumbers which may reduce the performance of techniques tailored to classify samples into classes, i.e., authentic or fraudulent. This paper proposes a novel method for selecting subsets of wavenumbers (variables) that better classify samples into such classes. For that matter, principal components analysis (PCA) is integrated to the k-nearest neighbor (KNN) classification technique. PCA is applied to FTIR-ATR data, and a variable importance index is built on the PCA outputs. An iterative backward variable elimination is started guided by that index; after each variable removal, samples are categorized into authentic or fraudulent classes using KNN, and the classification accuracy is measured. The wavenumber subset compromising high accuracy and reduced percent of retained variables is chosen. When applied to Cialis FTIR-ATR data, the proposed approach retained only average 1.84% of the original variables and increased the classification accuracy average 2.1%, to 0.9897 from 0.9689; as for Viagra data, the method increased average classification accuracy 1.56%, from 0.9135 to 0.9278, using only 7.72% of the original variables.

Data mining-based method for identifying discriminant attributes in sensory profiling

Selection of attributes from a group of candidates to be assessed through sensory analysis is an important issue when planning sensory panels. In attribute selection it is desirable to reduce the list of those to be presented to panelists to avoid fatigue, minimize costs and save time. In some applications the goal is to keep attributes that are relevant and non-redundant in the sensory characterization of products. In this paper, however, we are interested in keeping attributes that best discriminate between products. For that we present a data mining-based method for attribute selection in descriptive sensory panels, such as those used in the Quantitative Descriptive Analysis. The proposed method is implemented using Principal Component Analysis and the k-Nearest Neighbor classification technique, in conjunction with Pareto Optimal analysis. Objectives are ( i) to identity the set of attributes that best discriminate samples analyzed in the panel, and ( ii) to indicate the group of panelists that provide consistent evaluations. The method is illustrated through a case study where beef cubes in stew, used as combat ration by the American Army, are characterized in sensory panels using the Spectrum protocol.

Automated MS-Lesion Segmentation by K-Nearest Neighbor Classification

This paper proposes a new method for fully automated multiple sclerosis (MS) lesion segmentation in cranial magnetic resonance (MR) imaging. The algorithm uses the T1-weighted and the fluid attenuation inversion recovery scans. It is based the K-Nearest Neighbor (KNN) classification technique. The data has been acquired at the Children�s Hospital Boston (CHB) and the University of North Carolina (UNC). Manual segmentations, composed by a human expert of the CHB, were used for training of the KNN-classification. The method uses voxel location and signal intensity information for determination of the probability being a lesion per voxel, thus generating probabilistic segmentation images. By applying a threshold on the probabilistic images binary segmentations are derived. Automatic segmentations were performed on a set of testing images, and compared with manual segmentations from a CHB and a UNC expert rater. Furthermore, a combined segmentation was composed from segmentations from different algorithms, and used for evaluation. The proposed method shows good resemblance with the segmentations of the CHB rater. High specificity and lower specificity has been observed in comparison with the combined segmentations. Over- and undersegmentation can be easily corrected in this procedure by varying the threshold on the probabilistic segmentation image. The proposed method offers an automated and fully reproducible approach that accurate and applicable on standard clinical MR images.

Learning-aided predictor integration for system performance prediction

The integration of multiple predictors promises higher prediction accuracy than the accuracy that can be obtained with a single predictor. The challenge is how to select the best predictor at any given moment. Traditionally, multiple predictors are run in parallel and the one that generates the best result is selected for prediction. In this paper, we propose a novel approach for predictor integration based on the learning of historical predictions. Compared with the traditional approach, it does not require running all the predictors simultaneously. Instead, it uses classification algorithms such as k-Nearest Neighbor (k-NN) and Bayesian classification and dimension reduction technique such as Principal Component Analysis (PCA) to forecast the best predictor for the workload under study based on the learning of historical predictions. Then only the forecasted best predictor is run for prediction. Our experimental results show that it achieved 20.18% higher best predictor forecasting accuracy than the cumulative MSE based predictor selection approach used in the popular Network Weather Service system. In addition, it outperformed the observed most accurate single predictor in the pool for 44.23% of the performance traces.

An ensemble-driven k-NN approach to ill-posed classification problems

This paper addresses the supervised classification of remote-sensing images in problems characterized by relatively small-size training sets with respect to the input feature space and the number of classifier parameters (ill-posed classification problems). An ensemble-driven approach based on the k-nearest neighbor ( k-NN) classification technique is proposed. This approach effectively exploits semilabeled samples (i.e., original unlabeled samples labeled by the classification process) to increase the accuracy of the classification process. Experimental results obtained on ill-posed classification problems confirm the effectiveness of the proposed approach, which significantly increases both the accuracy and the reliability of classification maps.

Probabilistic segmentation of brain tissue in MR imaging

A new method has been developed for probabilistic segmentation of five different types of brain structures: white matter, gray matter, cerebro-spinal fluid without ventricles, ventricles and white matter lesion in cranial MR imaging. The algorithm is based on information from T1-weighted (T1-w), inversion recovery (IR), proton density-weighted (PD), T2-weighted (T2-w) and fluid attenuation inversion recovery (FLAIR) scans. It uses the K-Nearest Neighbor classification technique that builds a feature space from spatial information and voxel intensities. The technique generates for each tissue type an image representing the probability per voxel being part of it. By application of thresholds on these probability maps, binary segmentations can be obtained. A similarity index (SI) and a probabilistic SI (PSI) were calculated for quantitative evaluation of the results. The influence of each image type on the performance was investigated by alternately leaving out one of the five scan types. This procedure showed that the incorporation of the T1-w, PD or T2-w did not significantly improve the segmentation results. Further investigation indicated that the combination of IR and FLAIR was optimal for segmentation of the five brain tissue types. Evaluation with respect to the gold standard showed that the SI-values for all tissues exceeded 0.8 and all PSI-values exceeded 0.7, implying an excellent agreement.

A new approach to mining fuzzy databases using nearest neighbor classification by exploiting attribute hierarchies

Data classification is a well-organized operation in the field of data mining. This article presents an application of the k-nearest neighbor classification technique for mining a fuzzy database. We consider a data set in which attribute values have certain similarities in nature and analyze the observations for the domain of each attribute, on the basis of fuzzy similarity relations. The proposed technique is general and the presented case study demonstrates the suitability of using this fuzzy approach for mining fuzzy databases, especially when the database contains various levels of abstraction. © 2004 Wiley Periodicals, Inc. Int J Int Syst 19: 1277–1290, 2004.

Automatic segmentation of different-sized white matter lesions by voxel probability estimation

A new method for fully automated segmentation of white matter lesions (WMLs) on cranial MR imaging is presented. The algorithm uses five types of regular MRI-scans. It is based on a K-Nearest Neighbor (KNN) classification technique, which builds a feature space from voxel intensity features and spatial information. The technique generates images representing the probability per voxel being part of a WML. By application of thresholds on these probability maps binary segmentations can be produced. ROC-curves show that the segmentations achieve a high sensitivity and specificity. Three similarity measures, the similarity index (SI), the overlap fraction (OF) and the extra fraction (EF), are calculated for evaluation of the results and determination of the optimal threshold on the probability map. Investigation of the relation between the total lesion volume and the similarity measures shows that the method performs well for lesions larger than 2 cc. The maximum SI per patient is correlated to the total WML volume. No significant relation between the lesion volume and the optimal threshold has been observed. The probabilistic equivalents of the SI, OF en EF (PSI, POF and PEF) allow direct evaluation of the probability maps, which provides a strong tool for comparison of different classification results. A significant correlation between the lesion volume and the PSI and the PEF has been noticed. This method for automated WML segmentation is applicable to lesions of different sizes and shapes, and reaches an accuracy that is comparable to existing methods for multiple sclerosis lesion segmentation. Furthermore, it is suitable for detection of WMLs in large and longitudinal population studies.

Probabilistic segmentation of white matter lesions in MR imaging

A new method has been developed for fully automated segmentation of white matter lesions (WMLs) in cranial MR imaging. The algorithm uses information from T1-weighted (T1-w), inversion recovery (IR), proton density-weighted (PD), T2-weighted (T2-w) and fluid attenuation inversion recovery (FLAIR) scans. It is based on the K-Nearest Neighbor (KNN) classification technique that builds a feature space from voxel intensities and spatial information. The technique generates images representing the probability per voxel being part of a WML. By application of thresholds on these probability maps, binary segmentations can be obtained. ROC curves show that the segmentations achieve both high sensitivity and specificity. A similarity index (SI), overlap fraction (OF) and extra fraction (EF) are calculated for additional quantitative analysis of the result. The SI is also used for determination of the optimal probability threshold for generation of the binary segmentation. Using probabilistic equivalents of the SI, OF and EF, the probability maps can be evaluated directly, providing a powerful tool for comparison of different classification results. This method for automated WML segmentation reaches an accuracy that is comparable to methods for multiple sclerosis (MS) lesion segmentation and is suitable for detection of WMLs in large and longitudinal population studies.