Data Classification Algorithms Research Articles

Multimodal Classification Technique for Fall Detection of Alzheimer’s Patients by Integration of a Novel Piezoelectric Crystal Accelerometer and Aluminum Gyroscope with Vision Data

Smart expert systems line up with various applications to enhance the quality of lifestyle of human beings, such as major applications for smart health monitoring systems. An intelligent assistive system is one such application to assist Alzheimer’s patients in carrying out day-to-day activities and real-time monitoring by the caretakers. Fall detection is one of the tasks of an assistive system; many existing methods primarily focus on either vision or sensor data. Vision-based methods suffer from false positive results because of occlusion, and sensor-based methods yield false results because of the patient’s long-term lying posture. We address this problem by proposing a multimodel fall detection system (MMFDS) with hybrid data, which includes both vision and sensor data. Random forest and long-term recurrent convolution networks (LRCN) are the primary classification algorithms for sensor data and vision data, respectively. MMFDS integrates sensor and vision data to enhance fall detection accuracy by incorporating an ensemble approach named majority voting for the hybrid data. On evaluating the proposed work on the UP fall detection dataset, accuracy was enhanced to 99.2%, with an improvement in precision, F1 score, and recall.

An automatic methodology to measure drivers’ behavior in public transport

The way in which public transport buses are driven has an influence in users’perception and satisfaction with the service. Bus driver’s behavior is usually obtained surveying passengers and/or using the mystery passenger method, not necessarily allowing for an objective and continuous evaluation. In this work, we introduce a novel methodology to automatically classify drivers’ behavior in a more consistent and objective manner, based on data from inertial measurement units, and machine learning techniques. By substituting human evaluators with automatic data collection and classification algorithms, we are able to reduce the subjectivity and cost of the current methodology, while increasing sample size. Our approach is based on three components: i) data capture using inertial measurement units (e.g. mobile devices), ii) carefully tuned classifiers that deal with sample imbalance problems, and iii) an interpretable scoring system. Results show that collected data captures several types of undesirable maneuvers, providing a rich information to the classification process. In terms of categorization performance, the evaluated classifiers, namely support vector machines, decision trees and k-NN, deliver high and consistent accuracy after the tuning process, even in the presence of a highly imbalanced sample. Finally, the proposed driver’s behavior score shows high discriminative power, effectively characterizing differences between drivers, and providing driver-tailored driving recommendations, that can be generated in specific spots, in order to improve passengers’ experience. The resulting methodology can be cost-effectively deployed at a large scale with good performance.

Using Classification Algorithms in Data Mining in Diagnosing Breast Cancer

Data mining is the process of extracting useful information from large-scale data in an understandable and logical way. According to the main machine learning techniques of data mining; classification and regression, association rules and cluster analysis. Classification and regression are known as predictive models, and clustering and association rules are known as descriptive models. In this study, the classification method was used. With this method, it is aimed to assign a data set to one of the previously determined different classes. The data set used in the study was obtained from the UCIrvine Machine Learning Repository database. The dataset named “Breast cancer” consists of breast cancer data consisting of 699 samples and 10 features collected by William H. at the University of Wisconsin hospital. The data content includes information about the characteristics of some cells analyzed in the detection of breast cancer, cell division, and whether they are benign or malignant. Upon completion of the study, a classification process is performed by determining whether the targeted person has cancerous or non-cancerous cells. In the study carried out in this context; Data mining analyzes were performed using WEKA and Orange programs, SVM (Support Vector Machine), Random Forest algorithms. Along with the analysis results, a comparison was made on the data set, taking into account the previous studies. It is aimed that the conclusions obtained at the end of the study will guide medical professionals working in this field in the diagnosis of breast cancer.

Conditional self-attention generative adversarial network with differential evolution algorithm for imbalanced data classification

Imbalanced data classification is an important research topic in real-world applications, like fault diagnosis in an aircraft manufacturing system. The over-sampling method is often used to solve this problem. It generates samples according to the distance between minority data. However, the traditional over-sampling method may change the original data distribution, which is harmful to the classification performance. In this paper, we propose a new method called Conditional Self-Attention Generative Adversarial Network with Differential Evolution (CSAGAN-DE) for imbalanced data classification. The new method aims at improving the classification performance of minority data by enhancing the quality of the generation of minority data. In CSAGAN-DE, the minority data are fed into the self-attention generative adversarial network to approximate the data distribution and create new data for the minority class. Then, the differential evolution algorithm is employed to automatically determine the number of generated minority data for achieving a satisfactory classification performance. Several experiments are conducted to evaluate the performance of the new CSAGAN-DE method. The results show that the new method can efficiently improve the classification performance compared with other related methods.

An improved deep learning algorithm in enabling load data classification for power system

Load behaviors significantly impact the planning, dispatching, and operation of the modern power systems. Load classification has been proved as one of the most effective ways of analyzing the load behaviors. However, due to the issues of data collection, transmission, and storage in current power systems, data missing problems frequently occur, which prevents the load classification tasks from precisely identifying the load classes. Simultaneously, because of the diversities of the load categories, different loads contribute various amounts of data, which causes the class imbalance issue. The traditional load data classification algorithms lack the ability to solve the aforementioned issues, which may deteriorate the load classification accuracy. Therefore, this study proposed an improved deep learning algorithm based on the load classification approach in terms of raising the classification performances with solving the data missing and class imbalance issues. First, the LATC (low-rank autoregressive tensor completion) algorithm is used to solve the data missing issue to improve the quality of the training dataset. A Borderline-SMOTE algorithm is further adopted to improve the class distribution in the training dataset to improve the training performances of biGRU (bidirectional gated recurrent unit). Afterward, to improve the classification accuracy in the classification task, the biGRU algorithm, combined with the attention mechanism, is used as the underlying infrastructure. The experimental results show the effectiveness of the proposed approach.

Retraction Note: A novel large-scale multimedia image data classification algorithm based on mapping assisted deep neural network

Retraction Note: A novel large-scale multimedia image data classification algorithm based on mapping assisted deep neural network

Application of Improved DEA Algorithm in Public Management Problem Classification.

In order to effectively solve the related problems in the process of public management, this research proposes an improved algorithm technology based on Data Envelopment Analysis (DEA) and classification algorithm. With the help of lever management, this paper further solves and overcomes the problem of DEA algorithm itself and the defect of “relative effectiveness.” At the same time, in order to avoid the impact on the input and output indicators, with the help of principal component analysis, taking the performance evaluation of public management departments as the research direction, this paper makes an empirical analysis on the performance evaluation of public management departments. The evaluation results of the index system show that the correlation coefficient between the efficiency value of the initial index system and the efficiency value of optimization 2 is 0.977759, and the correlation coefficient is less than 0.7. The evaluation results are more reasonable than those before the improvement.

TSFNFR: Two-stage fuzzy neighborhood-based feature reduction with binary whale optimization algorithm for imbalanced data classification

Recently developed imbalanced data classification models are mainly focused on the majority class samples. In addition, several whale optimization-based feature reduction models are inefficient for high-dimensional data, readily fall into the local optimum, and are subject to difficulties associated with achieving a global optimal feature subset, due to high costs. To overcome the drawbacks, in this study, a two-stage feature reduction model using fuzzy neighborhood rough sets (FNRS) and the binary whale optimization algorithm (BWOA) was developed for imbalanced data classification. First, to indicate the sample fuzziness of mixed data, a similarity measure between samples based on fuzzy neighborhood was defined to investigate the similarity matrix and fuzzy neighborhood granule, and a new FNRS model was presented by constructing lower and upper approximations. By considering the uneven distribution of classes, the boundary-based feature significance measure was developed to minimize the influence of the uncertainty in boundary region for imbalanced data. Second, fuzzy neighborhood roughness and decision entropy were investigated based on FNRS, and by integrating these above measures, fuzzy neighborhood decision entropy was proposed to evaluate the fuzziness and roughness of the fuzzy neighborhood for imbalanced data. The external and internal significance metrics were proposed to achieve the preselected feature subset in the first stage. Third, in this second stage, a new control factor was defined to control the position of whales, and a novel fitness function was developed to evaluate the selected feature subset from imbalanced datasets. Thereafter, the immune regulation strategy of artificial immune was introduced into the BWOA to design the mixed selection probability, to divide the whale population. Two local interference strategies were applied to adjust the whale position and prevent BWOA trapped in the local optimum. Thus, an optimal feature subset was achieved by constantly iterating the BWOA. Finally, a two-stage feature reduction algorithm was designed to handle imbalanced and high-dimensional data, where the particle swarm optimization (PSO) algorithm was employed to determine the different optimized parameters for this two-stage algorithm. Experiments conducted on 22 datasets revealed that the proposed algorithm is efficient for two-class and multiclass datasets.

Construction of Unbalanced Engineering Cost Management and Multivariate Statistics System Based on Random Matrix Weight Algorithm

The informatization of engineering cost plays an important role in the cost management and is also one of the important subjects of the informatization in the engineering field. It is the direction of the future project cost management. There is a trend of exponential increase in the information about project cost, which brings a great challenge to cost informatization. On the basis of the construction quantity and cost composition in the process of analyzing the present situation of the project cost, the overall index of the construction quantity and cost is constructed. For bridges and tunnels and to achieve a reasonable index of roadbed, on the basis of considering project at different stages and the demand for different levels of cost management personnel, adopt the method of statistical analysis and neural network, respectively, and set up the project cost fast estimation method and cost prediction model: intelligent prediction model and index system forecasting model. The error analysis and checking calculation of the two models provide the possibility for correctly estimating the engineering quantity and engineering cost in the preliminary design stage. Aiming at the low accuracy of the traditional unbalanced cost data classification algorithm, the dimension of random matrix weight was introduced into the cost data classification. On the basis of analyzing the spatial and temporal characteristics of WAMS measurement data, a high-dimensional random matrix model of WAMS measurement big data was constructed according to the high-dimensional random matrix theory, and a nonequilibrium cost data classification model based on the weighted algorithm of improved fuzzy rules was established. The model is applied to a unit project cost data classification, which verifies the operability of the model and provides a new idea for other units to carry out similar work. Finally, other common problems that should be paid attention to when using multivariate statistical analysis method are put forward.

Design and Implementation of Financial Service and Management Platform considering Support Vector Machine Algorithm.

With the rapid economic development, the financial industry has quietly become the leader of industries, the core and lifeblood of promoting economic development. At the same time, various financial services and management platforms emerge one after another. However, the emergence of financial services and management platforms cannot effectively alleviate the current financial crisis. In the face of increasingly complex financial risks, traditional financial service and management platforms cannot achieve effective information sharing, which leads to continued low service and management efficiency and frequent financial risk problems. Support vector machine is a data classification algorithm based on supervision, which can realize data sharing and improve the efficiency of data processing. The article firstly readjusted the underlying architecture of the financial service and management platform to break through the barriers of data interaction. Then on this basis, the article further combines the support vector machine algorithm and extends it from binary data classification to multivariate classification. Finally, the paper redesigns the financial service and management platform considering support vector machines. After a series of experiments, it can be found that the financial service and management platform based on the support vector machine algorithm can reduce the financial risk by 17.2%, improve the financial service level by 30.2%, and improve the financial comprehensive service level by 45.2%. At the same time, thanks to information sharing and interaction, the financial service and management platform can effectively predict financial risks, and the accuracy of the prediction basically reaches 78.9%. This shows that a financial service and management platform that takes into account the support vector machine algorithm can effectively prevent financial risks and improve the efficiency of financial services and management.

Многопризнаковая петрофизическая классификация горных пород как основа интерпретации геофизических данных

The work is carried out approbation of the algorithm for the complex interpretation of geophysical data using data classification algorithm. Distributions by class are obtained and compared with petrophysical materials from published sources. This data is compared and geological composition of the studied area is obtained. The applicability of this approach for the integrated interpretation of geophysical data is shown, including low-level studied areas with a complex geological structure.

Improving medical diagnostics with machine learning: a study on data classification algorithms

Improving medical diagnostics with machine learning: a study on data classification algorithms

Fault Diagnosis of Oil‐Immersed Power Transformers Using SVM and Logarithmic Arctangent Transform

A new method of dissolved gas analysis is proposed to improve the accuracy of transformer fault diagnosis. The slime mold optimized support vector machine (SMA-SVM), and logarithmic arctangent transform (LOG-ACT) are combined. On the one hand, the better global optimization performance of SMA is used to optimize SVM parameters to solve the difficulty of SVM parameter selection. On the other hand, corresponding transformations are carried out for different features: the logarithmic(LOG) transformation is carried out for the original DGA data to retain the order of magnitude information. The arctangent (ACT) transformation is carried out for the ratio features to improve the data structure. Therefore, the combination of data transformation and optimization model can improve the accuracy of diagnosis from two aspects of data structure and classification algorithm. The performance of the proposed method was compared with IEC three ratio method, artificial neural network, optimized artificial neural network, GA-SVM, and PSO-SVM. Experimental results using published data show that the proposed method can significantly improve the accuracy of transformer fault diagnosis. © 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

Adoption of Dexmedetomidine in Different Doses at Different Timing in Perioperative Patients.

Dexmedetomidine (Dex) is an alpha-2 agonist used for sedation during various procedures. Dex activates 2-adrenoceptors, and causes the decrease of sympathetic tone, with attenuation of the neuroendocrine and hemodynamic responses to anesthesia and surgery; it reduces anesthetic and opioid requirements; and causes sedation and analgesia. Objective: it was to compare the perioperative effects of different doses of Dex at different timing in patients using Dex during the perioperative period adopting a medical data classification algorithm based on optimized semi-supervised collaborative training (Tri-training). Methods: 495 patients requiring surgical treatment in Xingtai People's Hospital were randomly selected as the study subjects. The patients were divided into group A (used before induction), group B (used during induction), and group C (used after induction) according to different induction timing, with 165 cases in each group. Then, groups A, B, and C were divided into groups A1, B1, and C1 (0.4 μg/(kg·h) rate), groups A2, B2, and C2 (0.6 μg/(kg·h) rate), and groups A3, B3, and C3 (0.8 μg/(kg·h) rate) according to the dose used, with 55 cases in each group. Intraoperative anesthesia and postoperative adverse reactions were compared among the 9 groups. Results: the similarity between the Tri-training algorithm optimized by Naive Bayes (NB) classification algorithm and the actual classification (93.49%) was clearly higher than that by decision tree (DT) and K-nearest neighbor (kNN) classification algorithm (76.21%, 74.31%); heart rate (HR), systolic blood pressure (SBP), and diastolic blood pressure (DBP) values decreased obviously after Dex used in groups B1, C1, B2, C2, B3, and C3 (P <0.05), but did not change significantly in groups A1, A2, and A3 (P >0.05); the proportion of patients with satisfactory Ramsay score in group A3 was distinctly superior than that in groups A1, A2, B1, B2, B3, C1, C2, and C3 (P <0.05); the incidence of adverse reactions in group A3 was significantly inferior than that in groups A1, A2, B1, B2, B3, C1, C2, and C3 (P <0.05). Conclusion: the optimization effect of NB classification algorithm was the best, and the injection of Dex at the injection rate of 0.8 μg/(kg·h) before induction of anesthesia could apparently improve the fluctuation of HR, SBP, and DBP during perioperative period, and effectively reduce the occurrence of adverse reactions in patients, with better sedative effect on patients.

Hybrid Optimization-Based Structural Design of Deep Q Network With Feature Selection Algorithm for Medical Data Classification

In the area of medical informatics, the medical data classification is considered a complicated job. However, accurate classification of medical data is a complex task. Therefore, a robust and effective hybrid optimization-based deep learning method for classifying the medical data is developed in this research. The input data is pre-processed using data normalization method. Then, the features are selected using the proposed Henry Sea Lion Optimization (HSLnO), which is the combination of Henry Gas Solubility Optimization (HGSO) and Sea Lion Optimization (SLnO). The classification process is achieved using an optimized Deep Q Network (DQN). The DQN is optimized using the proposed Shuffled Shepherd Whale optimization Algorithm (SSWOA). The proposed SSWOA is developed by the integration of the Shuffled Shepherd Optimization Algorithm (SSOA) and Whale Optimization Algorithm (WOA). The developed technique achieves better performance of testing accuracy, sensitivity, and specificity with values of 95.413%, 95.645%, and 95.364%, respectively.

Two-step learning for crowdsourcing data classification.

Crowdsourcing learning (Bonald and Combes 2016; Dawid and Skene, J R Stat Soc: Series C (Appl Stat), 28(1):20–28 1979; Karger et al. 2011; Li et al, IEEE Trans Knowl Data Eng, 28(9):2296–2319 2016; Liu et al. 2012; Schlagwein and Bjorn-Andersen, J Assoc Inform Syst, 15(11):3 2014; Zhang et al. 2014) plays an increasingly important role in the era of big data (Liu et al., IEEE Trans Syst Man Cybern: Syst, 48(12): 451–2461, 2017; Zhang et al. 2014) due to its ability to easily solve large-scale data annotations (Musen et al., J Amer Med Informs Assoc, 22(6):1148–1152 2015). However, in the process of crowdsourcing learning, the uneven knowledge level of workers often leads to low accuracy of the label after marking, which brings difficulties to the subsequent processing (Edwards and Teddy 2013) and analysis of crowdsourcing data. In order to solve this problem, this paper proposes a two-step learning crowdsourced data classification algorithm, which optimizes the original label data by simultaneously considering the two issues of different worker abilities and the similarity between crowdsourced data (Kasikci et al. 2013) samples, so as to get more accurate label data. The two-step learning algorithm mainly includes two steps. Firstly, the worker’s ability to label different samples is obtained by constructing and training the worker’s ability model, and then the similarity between samples is calculated by the cosine measurement method (Muflikhah and Baharudin 2009), and finally the original label data is optimized by combining the above two results. The experimental results also show that the two-step learning classification algorithm proposed in this article has achieved better experimental results than the comparison algorithm.

Optimising Multilayer Perceptron weights and biases through a Cellular Genetic Algorithm for medical data classification

In recent years, technology in medicine has shown a significant advance due to artificial intelligence becoming a framework to make accurate medical diagnoses. Models like Multilayer Perceptrons (MLPs) can detect implicit patterns in data, allowing identifying patients conditions that cannot be seen easily. MLPs consist of biased neurons arranged in layers, connected by weighted connections. Their effectiveness depends on finding the optimal weights and biases that reduce the classification error, which is usually done by using the Back Propagation algorithm (BP). But BP has several disadvantages that could provoke the MLP not to learn. Metaheuristics are alternatives to BP that reach high-quality solutions without using many computational resources. In this work, the Cellular Genetic Algorithm (CGA) with a specially designed crossover operator called Damped Crossover (DX), is proposed to optimise weights and biases of the MLP to classify medical data. When compared against state-of-the-art algorithms, the CGA configured with DX obtained the minimal Mean Square Error value in three out of the five considered medical datasets and was the quickest algorithm with four datasets, showing a better balance between time consumed and optimisation performance. Additionally, it is competitive in enhancing classification quality, reaching the best accuracy with two datasets and the second-best accuracy with two of the remaining. • The Cellular Genetic Algorithm is used to optimise a Feed Forward Neural Network. • A new crossover operator, called the Damped Crossover, is proposed. • Five benchmark medical datasets were considered for experiments. • Comparisons are made against nine state-of-the-art algorithms. • This proposal proves to be able to improve the classification power.

A novel oversampling and feature selection hybrid algorithm for imbalanced data classification

A novel oversampling and feature selection hybrid algorithm for imbalanced data classification

An Improved TCN Considering Data Augmentation in Enabling Load Classification

In modern power system, along with the developments of the data collecting technologies, the intensive and high-dimensional load data collection can be achieved. Therefore, to deeply reveal the patterns and behaviors hidden in the load dataset using load classification is of great significance for improving the service quality and the user experience of the power system. However, inevitable issues, for example the data missing and class imbalance are frequently reported in the present load dataset, which deteriorates the performance of the classification algorithms. Also, due to the special features, for example the time series, periodicity, and fluctuation of the load data, the traditional data classification algorithms also encounter performance defects. Therefore, this paper presents a data augmentation based enhanced temporal convolutional network (TCN) algorithm in enabling load classification. In the data augmentation phase, first an LRTC-TSVD algorithm is presented to implement the missing data completion. Second, a WGAN based class balancing approach is further presented to solve the class imbalance issue. Then, in the enhanced TCN phase, a WeightNorm, exponential linear unit (ELU) activation function, residual connection, and bidirectional feature fusion techniques based improved TCN (ITCN) algorithm is presented to carry out the accurate load data classification. Combining the data augmentation and the enhanced TCN phases, the ITCN algorithm is finally conducted. Based on the benchmark load datasets, the performances of the presented ITCN are evaluated. The experimental results report that the presented data augmentations can improve the quality of the dataset, moreover the classification algorithm is able to achieve the satisfied classification accuracy.

An Innovative Approach for Personnel Positioning

The contribution of the paper is the building of a model to predict the proper employees’ allocation in the Greek public sector. To acknowledge the set and weights of criteria upon which our model feeds data, a validated questionnaire was developed and used to conduct a primary quantitative survey amongst HR departments and employees of state organizations in Greece. On the acquired findings, several experiments were administered using linear and machine learning tools, aiming to replace time consuming and subjective procedures, followed by many organizations. Concluding, data classification algorithms are proposed to predict the best matching of employees, giving as inputs personnel qualifications as well as job specifications, leading to a model based on J48, a decision tree algorithm.