Nearest-neighbor Classifier Research Articles

An Evaluation of Fuzzy in Image Enhancement: Design and Comparison for Penicillium and Aspergillus Species

The main focus in this study is to enhance and classify the image of a type of filamentous fungi named Penicillium and Aspergillus. For image enhancement, fuzzy-partition gamma adaptive histogram equalization (FpGAHE) is proposed to improve the quality of an image, in particular the low quality of a microscopic image. Two stages have been considered in this technique. In the first stage, a fuzzy partition is developed to handle the inconsistency of the grey level values of the images by introducing a fuzzy set. In the second stage, surrounding neighbourhood is employed to avoid the imbalance data and reduce the drastically changes of brightness values of the image. The performances are evaluated into two parts i.e., image processing and image classification by using the collected microscopic images of fungi species. To evaluate the effectiveness of the proposed technique, the existing techniques, HE, AHE, CLAHE, GC and AGC is compared to the FpGAHE. In image processing, the result attained shows that the proposed technique has a better performance by obtaining the highest value for the PSNR, SSIM and FSIM evaluation for the species of A. terreus in clean condition. Meanwhile, in image classification, five different nearest neighbour classifiers have been tested. The results show the proposed FpGAHE with Improved Fuzzy-Based k Nearest Centroid Neighbour (IFkNCN) classifier perform the best result compare to other nearest neighbour classifier by obtaining the value of 92.59 and 93.95 for the salt and pepper and Gaussian noise corrupted images respectively.

A distributed evolutionary based instance selection algorithm for big data using Apache Spark

Instance selection is an important preprocessing technology in data mining and machine learning. In this paper, we proposed a novel evolutionary based instance selection algorithm for big data. First, we defined a coarse granularity chromosome structure to reduce the size of search space and costs of chromosome operations(recombination and mutation, etc.). Then a stratified evolution strategy was proposed to remove the hyper parameter in classic fitness function and achieve precise control over the reduction ratio of instances. Finally, a sampling-based fitness function was proposed to reduce the time complexity. Experimental results shown that our new algorithm is efficient to complete the instance selection task on data set with millions of instances in minutes-level. The 10-fold cross-validation also proved that the selection results on many datasets have high nearest neighbor classification accuracy.

Research on the Recommendation System of Music e-learning Resources with Blockchain based on Hybrid Deep Learning Model

Learners are confronted with an ever-growing array of diverse and complex educational resources as music education increasingly moves to online platforms. Traditional resource curation methods, which rely heavily on educators, fall short of meeting the dynamic needs of modern students. To address this issue, we present a novel recommendation system for music e-learning resources that combines the power of blockchain technology with a hybrid deep learning model. Our model combines blockchain's robust security and transparency features with advanced deep learning algorithms, enhancing the personalization and efficiency of resource recommendations. A backpropagation neural network with K nearest neighbor classification, traditional collaborative filtering (CF), and an improved CF algorithm are used in the hybrid approach. For the back propagation neural network algorithm, K nearest neighbor classification algorithm, traditional collaborative filtering (CF) and improved CF algorithm, the accuracy rate of improved CF algorithm is higher, reaching 95%. Comparing the proposed model with the association rule-based recommendation model and the content-based recommendation model, the model constructed in this study received high evaluation from experts, with an average score of 98, and more than 97% of them gave a high score of 95 or more, and the evaluation of experts tended to be consistent. Overall, the model proposed in this study can make better recommendations for music education learning resources and bring users a good learning experience, so this study has some practical application value. This research demonstrates a highly effective, blockchain-enhanced recommendation system for music e-learning resources. Our model has significant practical value and potential for adoption in online music education platforms because it provides tailored educational content and an enhanced learning experience.

Enhancing Multi-Source Open-Set Domain Adaptation Through Nearest Neighbor Classification With Self-Supervised Vision Transformer

Enhancing Multi-Source Open-Set Domain Adaptation Through Nearest Neighbor Classification With Self-Supervised Vision Transformer

A Comparative Study of Machine Learning Classifiers for Enhancing Knee Osteoarthritis Diagnosis

Knee osteoarthritis (KOA) is a leading cause of disability, particularly affecting older adults due to the deterioration of articular cartilage within the knee joint. This condition is characterized by pain, stiffness, and impaired movement, posing a significant challenge in medical diagnostics and treatment planning, especially due to the current inability for early and accurate detection or monitoring of disease progression. This research introduces a multifaceted approach employing feature extraction and machine learning (ML) to improve the accuracy of diagnosing and classifying KOA stages from radiographic images. Utilizing a dataset of 3154 knee X-ray images, this study implemented feature extraction methods such as Histogram of Oriented Gradients (HOG) with Linear Discriminant Analysis (LDA) and Min–Max scaling to prepare the data for classification. The study evaluates six ML classifiers—K Nearest Neighbors classifier, Support Vector Machine (SVM), Gaussian Naive Bayes, Decision Tree, Random Forest, and XGBoost—optimized via GridSearchCV for hyperparameter tuning within a 10-fold Stratified K-Fold cross-validation framework. An ensemble model has also been made for the already high-accuracy models to explore the possibility of enhancing the accuracy and reducing the risk of overfitting. The XGBoost classifier and the ensemble model emerged as the most efficient for multiclass classification, with an accuracy of 98.90%, distinguishing between healthy and unhealthy knees. These results underscore the potential of integrating advanced ML methodologies for the nuanced and accurate diagnosis and classification of KOA, offering new avenues for clinical application and future research in medical imaging diagnostics.

A novel machine learning algorithm selects proteome signature to specifically identify cancer exosomes

Non-invasive early cancer diagnosis remains challenging due to the low sensitivity and specificity of current diagnostic approaches. Exosomes are membrane-bound nanovesicles secreted by all cells that contain DNA, RNA, and proteins that are representative of the parent cells. This property, along with the abundance of exosomes in biological fluids makes them compelling candidates as biomarkers. However, a rapid and flexible exosome-based diagnostic method to distinguish human cancers across cancer types in diverse biological fluids is yet to be defined. Here, we describe a novel machine learning-based computational method to distinguish cancers using a panel of proteins associated with exosomes. Employing datasets of exosome proteins from human cell lines, tissue, plasma, serum, and urine samples from a variety of cancers, we identify Clathrin Heavy Chain (CLTC), Ezrin, (EZR), Talin-1 (TLN1), Adenylyl cyclase-associated protein 1 (CAP1), and Moesin (MSN) as highly abundant universal biomarkers for exosomes and define three panels of pan-cancer exosome proteins that distinguish cancer exosomes from other exosomes and aid in classifying cancer subtypes employing random forest models. All the models using proteins from plasma, serum, or urine-derived exosomes yield AUROC scores higher than 0.91 and demonstrate superior performance compared to Support Vector Machine, K Nearest Neighbor Classifier and Gaussian Naive Bayes. This study provides a reliable protein biomarker signature associated with cancer exosomes with scalable machine learning capability for a sensitive and specific non-invasive method of cancer diagnosis.

A novel machine learning algorithm selects proteome signature to specifically identify cancer exosomes.

Non-invasive early cancer diagnosis remains challenging due to the low sensitivity and specificity of current diagnostic approaches. Exosomes are membrane-bound nanovesicles secreted by all cells that contain DNA, RNA, and proteins that are representative of the parent cells. This property, along with the abundance of exosomes in biological fluids makes them compelling candidates as biomarkers. However, a rapid and flexible exosome-based diagnostic method to distinguish human cancers across cancer types in diverse biological fluids is yet to be defined. Here, we describe a novel machine learning-based computational method to distinguish cancers using a panel of proteins associated with exosomes. Employing datasets of exosome proteins from human cell lines, tissue, plasma, serum, and urine samples from a variety of cancers, we identify Clathrin Heavy Chain (CLTC), Ezrin, (EZR), Talin-1 (TLN1), Adenylyl cyclase-associated protein 1 (CAP1), and Moesin (MSN) as highly abundant universal biomarkers for exosomes and define three panels of pan-cancer exosome proteins that distinguish cancer exosomes from other exosomes and aid in classifying cancer subtypes employing random forest models. All the models using proteins from plasma, serum, or urine-derived exosomes yield AUROC scores higher than 0.91 and demonstrate superior performance compared to Support Vector Machine, K Nearest Neighbor Classifier and Gaussian Naive Bayes. This study provides a reliable protein biomarker signature associated with cancer exosomes with scalable machine learning capability for a sensitive and specific non-invasive method of cancer diagnosis.

An Investigation on the Use of Clustering Algorithms for Data Preprocessing in Breast Cancer Diagnosis

Classification algorithms are commonly used as a decision support system for diagnosing many diseases, such as breast cancer. The accuracy of classification algorithms can be affected negatively if the data contains outliers and/or noisy data. For this reason, outlier detection methods are frequently used in this field. In this study, we propose and compare various models that use clustering algorithms to detect outliers in the data preprocessing stage of classification to investigate their effects on classification accuracy. Clustering algorithms such as DBSCAN, HDBSCAN, OPTICS, FuzzyCMeans, and MCMSTClustering (MCMST) were used separately in the data preprocessing stage of the k Nearest Neighbor (kNN) classification algorithm for outlier elimination, and then the results were compared. According to the obtained results, MCMST algorithm was more successful in outlier elimination. The classification accuracy of the kNN + MCMST model was 0.9834, which was the best one, while the accuracy of kNN algorithm without using any data preprocessing was 0.9719.

A feature vector with insensitivity to the position of the outer race defect and its application in rolling bearing fault diagnosis

The fault diagnosis of rolling bearings is very important in industrial applications, which can avoid accidents and reduce operation and maintenance costs. Although the position of the bearing outer race defect has a significant impact on rolling bearing vibration response, most existing intelligent bearing fault diagnosis methods do not take this into account. In this paper, we establish a dynamic model of rolling bearing to clarify the influence of the outer race defect position on the dynamic response, and propose a feature vector that is insensitive to the outer race defect positions. First, the vibration characteristics of the outer race faults with different defect positions are analyzed, and the impact is evaluated using six indicators. Second, three indicators of insensitivity to the bearing outer race defect positions are constructed as the feature vector for bearing fault diagnosis. Finally, a bearing fault diagnosis method considering the positions of outer race defect is proposed based on the constructed feature vector and K nearest neighbor classifier. The diagnosis results of three datasets formed by experimental signals show that the constructed feature vector can separate different bearing states. Compared with the existing two diagnosis methods, the proposed diagnosis method obtains higher recognition accuracy, in the case of different outer race defect positions of the training set and the testing set. The above research results are expected to provide a reference for rolling bearing fault diagnosis, especially when considering the influence of the outer race defect positions.

Advancing Handwritten Musical Notation Recognition Using Deep Learning: A Convolutional Neural Network-Based Approach with Improved Accuracy

The use of computers to read musical scores is referred to as optical music recognition (OMR). The recent advancements in artificial intelligence and big data have led to the development of deep learning approaches for recognizing musical notes. Previous research has shown that there is a lot of room for improvement in handwritten musical notation recognition systems due to differences in writing styles and the complex structure of musical symbols. The research described here aims to develop a deep learning-based system for recognizing handwritten musical notation. The system uses a convolutional neural network (CNN) to extract and learn pixel features of musical symbols and achieve a recognition accuracy of over 90%. The CNN model was trained using image samples from the HOMUS dataset and fine-tuned to minimize the loss function and reduce classification errors. The CNN model achieved an accuracy of 96.95% on the test samples, which is a significant improvement over the 86.0% accuracy from previous studies. The performance of the CNN model was also compared to five state-of-the-art deep learning methods, namely, quantum gray Wolf optimization (QGWO) algorithm, nonfully connected network (NFC-Net) classifier, nearest neighbor classifier, data augmentation and ensemble learning, and the CNN model outperformed four of them. However, the CNN model occasionally misclassified musical symbols with similar shapes, indicating that there is still room for improvement in the system’s performance. Future research could focus on improving the model’s performance on similar-shaped symbols. Overall, the research demonstrates the effectiveness of using a CNN model for handwritten musical notation recognition and highlights the potential of deep learning approaches in this area.

Delicately Reinforced k-Nearest Neighbor Classifier Combined With Expert Knowledge Applied to Abnormity Forecast in Electrolytic Cell.

As the profit and safety requirements become higher and higher, it is more and more necessary to realize an advanced intelligent analysis for abnormity forecast of the synthetical balance of material and energy (AF-SBME) on aluminum reduction cells (ARCs). Without loss of generality, AF-SBME belongs to classification problems. Its advanced intelligent analysis can be realized by high-performance data-driven classifiers. However, AF-SBME has some difficulties, including a high requirement for interpretability of data-driven classifiers, a small number, and decreasing-over-time correctness of training samples. In this article, based on a preferable data-driven classifier, which is called a reinforced k -nearest neighbor (R-KNN) classifier, a delicately R-KNN combined with expert knowledge (DR-KNN/CE) is proposed. It improves R-KNN in two ways, including using expert knowledge as external assistance and enhancing self-ability to mine and synthesize data knowledge. The related experiments on AF-SBME, where the relevant data are directly sampled from practical production, have demonstrated that the proposed DR-KNN/CE not only makes an effective improvement for R-KNN, but also has a more advanced performance compared with other existing high-performance data-driven classifiers.

Comparison of the effects of imputation methods for missing data in predictive modelling of cohort study datasets

BackgroundMissing data is frequently an inevitable issue in cohort studies and it can adversely affect the study's findings. We assess the effectiveness of eight frequently utilized statistical and machine learning (ML) imputation methods for dealing with missing data in predictive modelling of cohort study datasets. This evaluation is based on real data and predictive models for cardiovascular disease (CVD) risk.MethodsThe data is from a real-world cohort study in Xinjiang, China. It includes personal information, physical examination data, questionnaires, and laboratory biochemical results from 10,164 subjects with a total of 37 variables. Simple imputation (Simple), regression imputation (Regression), expectation-maximization(EM), multiple imputation (MICE) , K nearest neighbor classification (KNN), clustering imputation (Cluster), random forest (RF), and decision tree (Cart) were the chosen imputation methods. Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) are utilised to assess the performance of different methods for missing data imputation at a missing rate of 20%. The datasets processed with different missing data imputation methods were employed to construct a CVD risk prediction model utilizing the support vector machine (SVM). The predictive performance was then compared using the area under the curve (AUC).ResultsThe most effective imputation results were attained by KNN (MAE: 0.2032, RMSE: 0.7438, AUC: 0.730, CI: 0.719-0.741) and RF (MAE: 0.3944, RMSE: 1.4866, AUC: 0.777, CI: 0.769-0.785). The subsequent best performances were achieved by EM, Cart, and MICE, while Simple, Regression, and Cluster attained the worst performances. The CVD risk prediction model was constructed using the complete data (AUC:0.804, CI:0.796-0.812) in comparison with all other models with p<0.05.ConclusionKNN and RF exhibit superior performance and are more adept at imputing missing data in predictive modelling of cohort study datasets.

A lost-in-space star identification algorithm based on regularized pattern recognition

Accuracy in spacecraft attitude determination has been improved continuously in parallel to scientific payloads requiring higher accuracy. Star trackers aim to maintain attitude determination by matching observations to the reference stars. They provide very high accuracy at any point in space at cost of higher computational complexity. This study presents an algorithm of lost-in-space star identification with high accuracy and acceptable computational complexity. This task is achieved by means of a novel dictionary-based star matching method by applying nearest neighbor classification through a binary search of Euclidean distances and regularization sequentially, accompanied by use of unconventional feature vectors representing each observation individually. The proposed method yields an estimated direction of the boresight vector orthogonal to the camera plane with an estimated axial rotation. The Hipparcos catalog is used for database generation, and the sensor parameters are taken from the project SharjahSat-1. The performance is evaluated in consideration of noise including position, brightness and false stars. By means of complexity analysis of database size and average run time, it is shown that the proposed method achieves better accuracy than other methods and maintains very high robustness to position noise and brightness noise. A high level of accuracy is achieved with an amount of database size and average run time that are competitive in the recent literature. It is demonstrated that the levels of pointing accuracy offered by state-of-the-art ADCS models are achieved by the proposed method with the given probabilities.

An Evaluation of Fuzzy in Image Enhancement: Design and Comparison for Penicillium and Aspergillus Species

The main focus in this study is to enhance and classify the image of a type of filamentous fungi named Penicillium and Aspergillus. For image enhancement, fuzzy-partition gamma adaptive histogram equalization (FpGAHE) is proposed to improve the quality of an image, in particular the low quality of a microscopic image. Two stages have been considered in this technique. In the first stage, a fuzzy partition is developed to handle the inconsistency of the gray level values of the images by introducing a fuzzy set. In the second stage, surrounding neighborhood is employed to avoid the imbalance data and reduce the drastically changes of brightness values of the image. The performances are evaluated into two parts i.e., image processing and image classification by using the collected microscopic images of fungi species. To evaluate the effectiveness of the proposed technique, the existing techniques, AHE and AGC is compared to the FpGAHE. In image processing, the result attained shows that the proposed technique has a better performance by obtaining the highest value for the PSNR, SSIM and FSIM evaluation for the species of A. terreus in clean condition. Meanwhile, in image classification, five different nearest neighbor classifiers have been tested. The results show the proposed FpGAHE with Improved Fuzzy-Based k Nearest Centroid Neighbor (IFkNCN) classifier perform the best result compare to other nearest neighbor classifier by obtaining the value of 92.59 and 93.95 for the salt and pepper and Gaussian noise corrupted images respectively.

The detection of alcohol intoxication using electrooculography signals from smart glasses and machine learning techniques

The operation of a motor vehicle under the influence of alcohol poses a significant risk to the safety of the driver, passengers, and other road users. Electrooculographic (EOG) signal analysis can be used to understand the movements and behavior of the eyes while driving. In our study, we used smart glasses to collect EOG data from nine participants who used a driving simulator. Their level of alcoholic intoxication was simulated by drunk vision goggles at three different levels of inebriation (0, 1, 2, and 3‰ blood alcohol content). We used machine learning algorithms (decision trees, support vector machines, nearest-neighbor classifiers, boosted trees, bagged trees, subspace discriminant classifier, subspace k nearest-neighbor classifier, and RUSBoosted Trees) to analyze the data. The Bagged Trees achieved the highest accuracy of 79%. The most important features to detect simulated alcohol intoxication were the blink rate and the velocity of the saccade, a rapid simultaneous movement of both eyes in the same direction. Our study shows the potential of using smart glasses and machine learning for the automated detection of alcohol intoxication, even when alcohol consumption is simulated.

Classification of EMG signals with CNN features and voting ensemble classifier

Electromyography (EMG) signals are primarily used to control prosthetic hands. Classifying hand gestures efficiently with EMG signals presents numerous challenges. In addition to overcoming these challenges, a successful combination of feature extraction and classification approaches will improve classification accuracy. In the current work, convolutional neural network (CNN) features are used to reduce the redundancy problems associated with time and frequency domain features to improve classification accuracy. The features from the EMG signal are extracted using a CNN and are fed to the ‘k’ nearest neighbor (KNN) classifier with a different number of neighbors ( 1 N N , 3 N N , 5 N N , and 7 N N ) . It results in an ensemble of classifiers that are combined using a hard voting-based classifier. Based on the benchmark Ninapro DB4 database and CapgMyo database, the proposed framework obtained 91.3 % classification accuracy on CapgMyo and 89.5 % on Ninapro DB4.

Data Augmentation with Nearest Neighbor Classifier for Few-Shot Named Entity Recognition.

Few-shot learning (FSL) is a category of machine learning models that are designed with the intent of solving problems that have small amounts of labeled data available for training. FSL research progress in natural language processing (NLP), particularly within the medical domain, has been notably slow, primarily due to greater difficulties posed by domain-specific characteristics and data sparsity problems. We explored the use of novel methods for text representation and encoding combined with distance-based measures for improving FSL entity detection. In this paper, we propose a data augmentation method to incorporate semantic information from medical texts into the learning process and combine it with a nearest-neighbor classification strategy for predicting entities. Experiments performed on five biomedical text datasets demonstrate that our proposed approach often outperforms other approaches.

Plant Leaf Disease Identification and Classification using Deep Learning Technique

In countries like India, whose important occupation is agriculture, face a huge loss when the crops get affected by any type of disease. These diseases attack the crops in various stages and can destroy the entire production. Since most diseases are transmitted from one crop to another there is an essential requirement to detect the type of disease in the early stage so that farmers can take the required action to “save the crops” and production. However, detection of the kind of disease in a huge amount of crops is very difficult for farmers, and sometimes it becomes more difficult due to lake of knowledge about the disease since there are various kinds of diseases. The automated detection of crop disease with images has been done using many classification techniques, such as k- Nearest Neighbor Classifier, Probabilistic Neural Network, Genetic Algorithm, Support Vector Machine, and Main Component Analysis, Artificial Neural Network, and Fuzzy Logic. In this paper to improve detection capability, CNN has been used with an available database having sufficient knowledge of the disease. In this simulation, analysis of CNN-based leaf disease detection has been done for different values of learning rate and with different training algorithms. K-fold cross-validation has been used for the validation of the classifier. With this configuration, ~90% accuracy has been achieved

A Fast Granular-Ball-Based Density Peaks Clustering Algorithm for Large-Scale Data.

Density peaks clustering algorithm (DP) has difficulty in clustering large-scale data, because it requires the distance matrix to compute the density and δ -distance for each object, which has O(n2) time complexity. Granular ball (GB) is a coarse-grained representation of data. It is based on the fact that an object and its local neighbors have similar distribution and they have high possibility of belonging to the same class. It has been introduced into supervised learning by Xia et al. to improve the efficiency of supervised learning, such as support vector machine, k -nearest neighbor classification, rough set, etc. Inspired by the idea of GB, we introduce it into unsupervised learning for the first time and propose a GB-based DP algorithm, called GB-DP. First, it generates GBs from the original data with an unsupervised partitioning method. Then, it defines the density of GBs, instead of the density of objects, according to the centers, radius, and distances between its members and centers, without setting any parameters. After that, it computes the distance between the centers of GBs as the distance between GBs and defines the δ -distance of GBs. Finally, it uses GBs' density and δ -distance to plot the decision graph, employs DP algorithm to cluster them, and expands the clustering result to the original data. Since there is no need to calculate the distance between any two objects and the number of GBs is far less than the scale of a data, it greatly reduces the running time of DP algorithm. By comparing with k -means, ball k -means, DP, DPC-KNN-PCA, FastDPeak, and DLORE-DP, GB-DP can get similar or even better clustering results in much less running time without setting any parameters. The source code is available at https://github.com/DongdongCheng/GB-DP.

Software Defect Prediction Using Non-Dominated Sorting Genetic Algorithm and $k$-Nearest Neighbour Classifier

Background: Software Defect Prediction (SDP) is a vital step in software development. SDP aims to identify the most likely defect-prone modules before starting the testing phase, and it helps assign resources and reduces the cost of testing. Aim: Although many machine learning algorithms have been used to classify software modules based on static code metrics, the k-Nearest Neighbors (kNN) method does not greatly improve defect prediction because it requires careful set-up of multiple configuration parameters before it can be used. To address this issue, we used the Non-dominated Sorting Genetic Algorithm (NSGA-II) to optimize the parameters in the kNN classifier with favor to improve SDP accuracy. We used NSGA-II because the existing accuracy metrics often behave differently, making an opposite judgment in evaluating SDP models. This means that changing one parameter might improve one accuracy measure while it decreases the others. Method: The proposed NSGAII-kNN model was evaluated against the classical kNN model and state-of-the-art machine learning algorithms such as Support Vector Machine (SVM), Naïve Bayes (NB), and Random Forest (RF) classifiers. Results: Results indicate that the GA-optimized kNN model yields a higher Matthews Coefficient Correlation and higher balanced accuracy based on ten datasets. Conclusion: The paper concludes that integrating GA with kNN improved defect prediction when applied to large or small or large datasets.