Data Classification Research Articles

A Method for Predicting the Result of University Employment Education Transformation Based on Multi-Source Information Technology

In the data on the outcomes of higher education employment transformation, there is high correlation among multiple independent variables, leading to multicollinearity. If multicollinearity exists, it can complicate the establishment of predictive analysis models and the interpretation of relationships between variables for higher education employment transformation outcomes. The employment issues of college graduates have become a focal point of social concern. These issues not only reflect the employment status of college graduates but also indirectly indicate the quality of higher education and the level of talent cultivation. Therefore, a prediction method for the effectiveness of higher education employment transformation based on multi-source information technology is proposed. According to the theory of multi-source information technology, an employment data classification framework is constructed to integrate multi-source information. The matching operator function output values are calculated, weights are determined, and the final college student information collection results are output. The standard error between employment unit types and salary ranges is analyzed, clustering training samples are trained, and the distance from each employment information sample to the centroid is calculated to obtain preprocessing results of the transformation outcomes, thus avoiding the impact of multicollinearity on the results of higher education employment transformation. A model structure for analyzing factors influencing higher education employment is established, allowing for an understanding of the characteristics of different types of educational resources and enabling the prediction of the effectiveness of higher education employment transformation. Experimental results show that this method can accurately monitor the dynamic employment situation of university students. After applying this method, the standard error and probability error values are significantly reduced, proving to be significantly effective in predicting the outcomes of higher education employment transformation.

Wearable Activity Recognition for Advancing Elderly Care with Modified Transformer Model

Technological devices such as smartphones can be utilized in tracking various human activities or movements through built-in accelerometers and gyroscopes. Data obtained from these inertial sensors can be utilized in various applications to assist people, including healthcare, human-computer interaction, and sports. As a result, further developments in the effective classification of time series data through machine and deep learning is highly valued and actively pursued. In this study, the transformer model, a deep learning architecture designed for sequential data such as natural language processing (NLP), has been utilized for analysis of time-series motion readings from wearable accelerometers. The transformer model in this study has been refined by incorporating a Long Short-Term Memory (LSTM) recurrent neural network (RNN) architecture. By leveraging the HAR70+ dataset with a wide range of activities, the modified transformer model in this study obtained a best accuracy of 95.85%, demonstrating that it can match the performance of state-of-the-art wearable activity recognition methods using Deep Neural Networks (DNN) and LSTM. Hence, the findings presented in this study suggest the future relevance of improved transformer or deep learning models to enhance the quality of life for seniors.

Detection of serum alterations in polysubstance use patients by FT-Raman spectroscopy

Substance use disorders pose significant health risks and treatment challenges due to the diverse interactions between substances and their impact on physical and mental health. The chemical effects of multiple substance use on bodily fluids are not yet fully understood. Therefore, this study aimed to investigate the chemical changes induced by a combination of substances compared to a control group. Analysis of FT-Raman spectra revealed structural alterations in the amide III, I, and C = O functional groups of lipids in subjects treated with opioids, alcohol and cannabis (polysubstance group). These changes were evident in the form of peak shifts compared to the control group. Additionally, an imbalance in the amide-lipid ratio was observed, indicating perturbations in serum protein and lipid levels. Furthermore, a 2D plot of two-track two-dimensional correlation spectra (2T2D-COS) demonstrated a shift towards dominance of lipid vibrations in the polysubstance use groups, contrasting with the predominance of the amide fraction in the control group. This observation suggests distinct molecular changes induced by multiple substance use, potentially contributing to the pathophysiology of substance use disorders. Principal Component Analysis (PCA) was utilized to visualize the data structure and identify outliers. Subsequently, Partial Least Squares Discriminant Analysis (PLS-DA) was employed to classify the polysubstance use and control groups. The PLS-DA model demonstrated high classification accuracy, achieving 100.00 % in the training dataset and 94.74 % in the test dataset. Furthermore, receiver operating characteristic (ROC) analysis yielded perfect AUC values of 1.00 for both the training and test sets, underscoring the robustness of the classification model.This study highlights the quantitative and qualitative changes in serum protein and lipid levels induced by polysubstance use groups, as evidenced by FT-Raman spectroscopy. The findings underscore the importance of understanding the chemical effects of polysubstance use on bodily fluids for improved diagnosis and treatment of substance use disorders. Moreover, the successful classification of spectral data using machine learning techniques emphasizes the potential of these approaches in clinical applications for substance abuse monitoring and management.

Optimized biomedical entity relation extraction method with data augmentation and classification using GPT-4 and Gemini.

Despite numerous research efforts by teams participating in the BioCreative VIII Track 01 employing various techniques to achieve the high accuracy of biomedical relation tasks, the overall performance in this area still has substantial room for improvement. Large language models bring a new opportunity to improve the performance of existing techniques in natural language processing tasks. This paper presents our improved method for relation extraction, which involves integrating two renowned large language models: Gemini and GPT-4. Our new approach utilizes GPT-4 to generate augmented data for training, followed by an ensemble learning technique to combine the outputs of diverse models to create a more precise prediction. We then employ a method using Gemini responses as input to fine-tune the BioNLP-PubMed-Bert classification model, which leads to improved performance as measured by precision, recall, and F1 scores on the same test dataset used in the challenge evaluation. Database URL: https://biocreative.bioinformatics.udel.edu/tasks/biocreative-viii/track-1/.

Availability and requirement of feed and fodder in different districts of Karnataka

Agriculture and livestock sectors are crucial for economic and social equity in rural areas. Livestock sector growth and fodder availability are directly related. Enhancing profitability of livestock husbandry can only be possible by ensuring adequate fodder to the livestock. A study was therefore conducted to know the feed and fodder availability and requirement at various districts of Karnataka. Secondary data of land use classification, crop production (2015 to 2019) and livestock census were used to derive dry matter (DM) availability and requirement for each district. Primary data was collected to understand the village level situation. Results indicated that state has an average of only 68.61±54.41% dry matter availability. Out of 30 districts, six districts belonged to surplus (>100%) and 12 districts belonged to severely deficient DM category with the DM availability of less than 40%. Crop residues contributed highest (69.01%) to the total DM. Contribution of total greens from all sources was found to be more in severely deficient districts. Bagalkot district had highest and Bengaluru urban had lowest DM availability. These findings are pivotal for developing fodder plan for Karnataka.

Multi-Class Imbalanced Data Handling with Concept Drift in Fog Computing: A Taxonomy, Review, and Future Directions

A network of actual physical objects or “IoT components” linked to the internet and equipped with sensors, electronics, software, and network connectivity is known as the Internet of Things (IoT). This ability of the IoT components to gather and share data is made possible by this network connectivity. Many IoT devices are currently operating, which generate a lot of data. When these IoT devices started collecting data, the cloud was the only place to analyze, filter, pre-process, and aggregate it. However, when it comes to IoT, the cloud has restrictions regarding latency and a more centralized method of distributing programs. A new form of computing called Fog computing has been proposed to address the shortcomings of current cloud computing. In an IoT context, sensors regularly communicate signal information, and edge devices process the data obtained from these sensors using Fog computing. The sensors’ internal or external problems, security breaches, or the integration of heterogeneous equipment contribute to the imbalanced data, i.e., comparatively speaking, one class has more instances than the other. As a result of this data, the pattern extraction is imbalanced . Recent attempts have concentrated heavily on binary-class imbalanced concerns with exactly two classes. However, the classification of multi-class imbalanced data is an issue that needs to be fixed in Fog computing, even if it is widespread in other fields, including text categorization, human activity detection, and medical diagnosis. The study intends to deal with this problem. It presents a systematic, thorough, and in-depth comparative analysis of several binary-class and multi-class imbalanced data handling strategies for batch and streaming data in IoT networks and Fog computing. There are five major objectives in this study. First, reviewing the Fog computing concept. Second, outlining the optimization metric used in Fog computing. Third, focusing on binary and multi-class batch data handling for IoT networks and Fog computing. Fourth, reviewing and comparing the current imbalanced data handling methodologies for multi-class data streams. Fifth, explaining how to cope with the concept drift, including novel and recurring classes, targeted optimization measures, and evaluation tools. Finally, the best performance metrics and tools for concept drift, binary-class (batch and stream) data, and multi-class (batch and stream) data are highlighted.

Tradisi “Marhaban” dalam Perspektif Maqashid Syari’ah

This study aims to further understand how the “marhaban” tradition is carried out in Semurup, Kerinci Regency. The focus is to what extent the “marhaban” tradition meets or may conflict with the principles of maqashid sharia. This study is an empirical study with a descriptive qualitative approach. Through this field research, the author will try to explore primary data sourced from the relevant place, namely in Semurup, Kerinci Regency. Data collection techniques are carried out by interview and observation. The data analysis techniques used include data reduction and classification, data presentation, and conclusion. The results of the study show that the “marhaban” tradition is carried out by completing the Qur'an first, dhikr, and prayers, strengthening family relationships, and maintaining cultural identity by involving community members and there is no waste or unnecessary expenditure in its implementation to maintain balance and justice in the management of the property. Parents also get inner peace because they feel they have started their children's lives with the blessings and protection of Allah.

Mental illness detection through harvesting social media: a comprehensive literature review

Mental illness is a common disease that at its extremes leads to personal and societal suffering. A complicated multi-factorial disease, mental illness is influenced by a number of socioeconomic and clinical factors, including individual risk factors. Traditionally, approaches relying on personal interviews and filling out questionnaires have been employed to diagnose mental illness; however, these manual procedures have been found to be frequently prone to errors and unable to reliably identify individuals with mental illness. Fortunately, people with mental illnesses frequently express their ailments on social media, making it possible to more precisely identify mental disease by harvesting their social media posts. This study offers a thorough analysis of how to identify mental illnesses (more specifically, depression) from users’ social media data. Along with the explanation of data acquisition, preprocessing, feature extraction, and classification techniques, the most recent published literature is presented to give the readers a thorough understanding of the subject. Since, in the recent past, the majority of the relevant scientific community has focused on using machine learning (ML) and deep learning (DL) models to identify mental illness, so the review also focuses on these techniques and along with their detail, their critical analysis is presented. More than 100 DL, ML, and natural language processing (NLP) based models developed for mental illness in the recent past have been reviewed, and their technical contributions and strengths are discussed. There exist multiple review studies, however, discussing extensive recent literature along with the complete road map on how to design a mental illness detection system using social media data and ML and DL classification methods is limited. The review also includes detail on how a dataset may be acquired from social media platforms, how it is preprocessed, and features are extracted from it to employ for mental illness detection. Hence, we anticipate that this review will help readers learn more and give them a comprehensive road map for identifying mental illnesses using users’ social media data.

Design of Time Series Fuzzy Segmentation Algorithm for Communication Data Classification

In order to solve the problem that the imbalance of communication data sets leads to a significant increase in classification difficulty, a classification algorithm for fuzzy segmentation of time series is proposed. The principal component analysis method is used to obtain the eigenvector with the largest eigenvalue. The time series of data is established according to the interval number theory. The segmentation target of the communication data time series is characterized by the Langmuir distance measurement function between data and categories. The classification result of fuzzy segmentation is obtained based on the judgment relationship between the difference of fuzzy classification matrix and the closing condition. The experimental results show that the number of data in the three classification situations of the algorithm in this paper is always at the corresponding ideal level, with high accuracy and low error and failure.

Adam energy valley optimization-based routing and RF-Spinalnet enabled medical data classification in IoT

Adam energy valley optimization-based routing and RF-Spinalnet enabled medical data classification in IoT

Exploratory parallel hybrid sampling framework for imbalanced data classification

Current engineering application scenarios often face the challenge of imbalanced data, hybrid sampling is an effective method to deal with the imbalanced data classification issue, which can avoid the issues of overfitting and mistakenly deleting useful majority samples when using oversampling approach and undersampling approach alone. However, at present most of the hybrid sampling approaches are implemented serially, and the implementation of oversampling and undersampling approaches alone will cause mutual interference and influence between them. This study proposes a parallel hybrid sampling framework based on the idea of parallel engineering and theoretically analyzes its superiority. The experimental results show that when applied to five classification algorithms with three performance evaluation metrics,the proposed framework outperforms the two mainstream hybrid sampling frameworks. Moreover, the proposed framework can effectively reduce the time consumption of hybrid sampling process.

Financial Risk Early Warning Model Combining SMOTE and Random Forest for Internet Finance Companies

At present, there have been many achievements on enterprise financial risk early warning model, but relatively few early warning studies focusing on the Internet industry. Research shows that the model has stable recognition accuracy and good prediction performance. The improved SMOTE algorithm based on PCA can realize the equalization of unbalanced data sets and use random forest as a classifier to classify and predict geological data. Because the noise data in the original data set may cause the change of the data distribution after interpolation, it is proposed to combine the PCA algorithm and the SMOTE algorithm, first perform noise reduction and dimension reduction, and then perform data interpolation to improve the classification performance of imbalanced data sets. my country's Internet financial listed companies conduct experiments on research samples, algorithm can better improve the classification accuracy and provide new ideas for the classification and prediction of unbalanced data.

Pre-operative dual-time-point [18F]FET PET differentiates CDKN2A/B loss and PIK3CA mutation status in adult-type diffuse glioma: a single-center prospective study.

While [18F]FET PET plays a complementary role in glioma imaging, it needs to be more comprehensively understood for improved characterization of glioma prior to surgery given the evolving landscape of molecular neuropathology. Thus, we investigated the utility of pre-operative dual-time-point [18F]FET PET in correlation with next-generation sequencing (NGS) data in patients with adult-type diffuse glioma (ADG). Adult patients who were suspected to have primary glioma were prospectively recruited between June 2021 and January 2024. They underwent pre-operative dual-time-point static PET/CT at 20min (early) and 80min (delay) after [18F]FET injection. Semi-quantitative parameters of the hottest lesion (SUVmax) of tumour and the hottest lesion-to-normal brain ratio (TBRmax) were assessed from each summed image. Furthermore, the percentage changes (△) of SUVmax and TBRmax between two images were calculated. Histopathology of glioma was determined according to the 2021 WHO classification and NGS data. This study investigated a dozen genes in 76 patients, of whom 51 had isocitrate dehydrogenase (IDH)-wild-type glioblastoma, 13 had IDH-mutant astrocytoma, and 12 had IDH-mutant oligodendroglioma. Every tumour was [18F]FET-avid having TBRmax more than 1.6. Patients with CDKN2A/B loss had significantly higher values of SUVmax (5.7 ± 1.6 vs. 4.7 ± 1.3, p = 0.004; 5.0 ± 1.4 vs. 4.4 ± 1.2, p = 0.026) and TBRmax (6.5 ± 1.8 vs. 5.1 ± 1.7, p = 0.001; 5.3 ± 1.5 vs. 4.3 ± 1.3, p = 0.004) in both scans than patients without CDKN2A/B loss, even after adjustment for age, MRI enhancement, tumor grade and type of pathology. Furthermore, patients with PIK3CA mutation (16.2 ± 11.8 vs. 6.7 ± 11.6, p = 0.007) had significantly higher △SUVmax than patients without PIK3CA mutation, even after adjustment for age, MRI enhancement, tumor grade, and type of pathology. Among the dozen genes investigated in this prospective study in patients with ADG, we found out that CDKN2A/B loss and PIK3CA mutation status could be differentiated by pre-operative dual-time-point [18F]FET PET/CT.

Interactive transformer and CNN network for fusion classification of hyperspectral and LiDAR data

ABSTRACT The Transformer has become pivotal for the integrated analysis of multi-source remote-sensing (RS) data in Earth observation, particularly in applications such as the fusion classification of hyperspectral images (HSI) and Light Detection and Ranging (LiDAR) data. However, Transformers are often employed as effective feature extractors by adopting similar processing blocks for different modalities from multi-source sensors, overlooking differences in imaging principles and data characteristics. Moreover, in the feature extraction process across different sensor data, there is a lack of necessary cross-modal information interaction, leading to insufficient utilization of complementary information between different sensors and resulting in suboptimal fusion outcomes. In this paper, we propose an interactive Transformer and CNN network for the fusion classification of HSI and LiDAR data. Specifically, a heterogeneous three-branch network architecture is designed for HSI and LiDAR data, where Transformers and CNNs encapsulate global contextual spatial and spectral information for HSI and capture geometric elevation patterns for LiDAR data, respectively. Elevation-Spatial Interaction (ESI) and Spectral-Spatial Interaction (SSI) modules are then introduced for multi-stage feature interaction. ESI enables the CNN-Transformer network to focus on essential local elevation details while simultaneously modelling global contextual spatial information. SSI facilitates the Transformer-Transformer network to cyclically intertwine spectral and spatial information for long-range spectral-spatial feature fusion. Finally, the interacted elevation, spatial, and spectral features undergo the Gated Fusion module to achieve hierarchical fusion adaptively, resulting in an elevation-spatial-spectral representation. Experiments conducted on three benchmark HSI-LiDAR datasets demonstrate the effectiveness of our proposed approach.

Automated system for classifying images of video streams for timely detection of fires

ABSTRACT Early detection of a fire during its development stage is highly desired to minimize human and material losses. The improvement of the efficiency of these monitoring systems through the analysis of fire video data drawn from unmanned aerial vehicles is considered . Classification of video images when monitoring a fire event on the territory is used based on their segmentation into identical rectangular segments of a given size and then assigning them to one of three classes: Smoke, Flame, and Indifferent classes. The Walsh-Hadamard transform was used to form descriptors for the Weak classifiers. Descriptors were calculated for three Weak classifiers; first, the Walsh-Hadamard transform was calculated for the window of the entire segment, and its spectral coefficients were used for the first Weak classifier. Next, the descriptors were calculated in windows whose sizes were half and a quarter of the original window size. The classifier consisted of three independently-trained neural networks. A simple ensemble averaging block was used to combine the outputs of those networks. A special software code has been developed to classify video images. The code allows us to create a database of segments of the ‘Smoke’ and ‘Flame’ classes, determine the two-dimensional Walsh-Hadamard spectrum of video image segments, train fully connected neural networks, conduct exploratory analysis and evaluate the relevance of the calculated two-dimensional spectral coefficients. For validation we used real data from Сlosed Circuit Television cameras presented in open databases. The experimental studies on the classification of video data containing flames and smoke showed that the proposed method has an average smoke detection accuracy of 86% and an average flame detection accuracy of 89.5%. Errors II in smoke detection and flame detection had the averages of 13% and 4.5%, respectively.

Predictive Analysis of Vector-Borne Diseases through Tabular Classification of Epidemiological Data

Predictive Analysis of Vector-Borne Diseases through Tabular Classification of Epidemiological Data

Extraction of Features for Time Series Classification Using Noise Injection

Time series data often display complex, time-varying patterns, which pose significant challenges for effective classification due to data variability, noise, and imbalance. Traditional time series classification techniques frequently fall short in addressing these issues, leading to reduced generalization performance. Therefore, there is a need for innovative methodologies to enhance data diversity and quality. In this paper, we introduce a method for the extraction of features for time series classification using noise injection to address these challenges. By employing noise injection techniques for data augmentation, we enhance the diversity of the training data. Utilizing digital signal processing (DSP), we extract key frequency features from time series data through sampling, quantization, and Fourier transformation. This process enhances the quality of the training data, thereby maximizing the model’s generalization performance. We demonstrate the superiority of our proposed method by comparing it with existing time series classification models. Additionally, we validate the effectiveness of our approach through various experimental results, confirming that data augmentation and DSP techniques are potent tools in time series data classification. Ultimately, this research presents a robust methodology for time series data analysis and classification, with potential applications across a broad spectrum of data analysis problems.

A-182 Applications of artificial intelligence optical character recognition in laboratory clerical functions offsetting staffing shortages and error reduction

Abstract Background The clinical laboratory profession has been dealing with a decade long, and now critical retention and staffing problem. The average national turnover rate for laboratory technicians is 17%, and the highest turnover rate exhibited for phlebotomists is 36%. While turnover rates have an impact on a laboratory, or hospitals budget, more importantly they have a substantial impact on quality and error rates, particularly in the pre-analytical stage of testing; with studies showing up to 65.09% of errors impacting the pre-analytical phase. Methods In an effort to remedy this high error rate, a meta-analysis was performed on paper order entry and clerical pre-analytical error, in parallel with performance and accuracy data of advanced optical character recognition driven by artificial intelligence and machine learning (AI OCR). Results AI OCR studies demonstrated an untrained accuracy percentage, to include handwriting of 91.08%, and when trained utilizing both relational and hierarchal logic data bases accuracy increased to 98.4%. The rate at which AI OCR processed and classified medical information was 3-4 pages per second. It is estimated that 99.9%-100% accuracy can be achieved with additional form field training and mapping. The level of accuracy and efficiency for laboratory paper order-entry utilizing AI OCR can also be increased with the addition of natural language processing and data recognition algorithms when combined with logic databases containing information such as medical record numbers, ICD-10 codes, and laboratory test compendiums. Conclusions The performance of this technology demonstrates a strong application to reduce the burden of clerical paper order entry, while addressing the 30.8% dissatisfaction rate regarding manual entry processes among lab professionals surveyed. Moreover, AI OCR clerical data classification, extraction and entry processes augmented with human verification steps examined, demonstrated overall accuracy of 99.99%. The application of AI OCR in laboratory pre-analytical processes has the potential to improve test order accuracy and efficiency, ultimately impacting the quality of patient care and operational costs for healthcare facilities. When examined through the lens of staff retention, studies show that 80% of laboratory employees feel that a workplace with the necessary tools and technology positively impacts their job satisfaction. While organizations investing in technologies that drive both job efficiency and satisfaction see a 125% return on investment through productivity. Demonstrating that the application of sophisticated technologies that both increase productivity and decrease error rates have a substantial impact on job satisfaction and ultimately retention. Although laboratory clerical data and order entry processes were examined for this meta-analysis, the aggregate data compiled showed that the application of AI OCR in clerical processes could also have a substantial impact on the 25.9% annual turnover rate for hospitals as a whole in the United States.

Comparing Sentinel-2 and Landsat 9 for land use and land cover mapping assessment in the north of Congo Republic: a case study in Sangha region

ABSTRACT A study of land use and land cover (LULC) mapping using satellite data was conducted in the Congo Basin rainforest in the Sangha area of the northern Congo Republic, which is crucial for sustainable land management. However, the heterogeneous nature of LULC features makes data classification difficult in this area because persistent cloud cover is challenging for observing the land surface and tropical humid conditions characterizing the study area. This study compares and evaluates the potential of two of the most widely used and popular satellite image sources, the recently launched Landsat 9 and Sentinel-2, to generate land-use land cover (LULC) maps in the Sangha area using pixel-based supervised machine learning algorithms, Random Forest (RF), for classification. All procedures were performed using ArcGIS Pro. The potential to improve Landsat 9 performance was tested using the nearest resampling technique, and different bands were resampled from 30 × 30 to 10 × 10 m. Owing to the heterogeneous characteristics of the Sangha study area, six classes were defined: dense forest, open forest, wetland forest, water bodies, urban areas, and bare soils. To classify and compare Sentinel-2 and Landsat 9, the shortwave infrared (SWIR), Near Infrared (NIR), and Red, Green, Blue (RGB) bands were used. These two satellite images were compared to test the quality of their results, particularly for assessing the accuracy of the LULC classifications and identifying which dataset had the highest accuracy. Results show that overall accuracy was 93.80% for Sentinel-2 while it was 91.60% for Landsat 9, similarly, the Kappa coefficient was calculated 0.89 and 0.85 for Sentinel-2 and Landsat 9, respectively. Therefore, Sentinel-2 exhibits significant classification ability compared to the nearest resample technique Landsat 9, despite the improved resolution of the latter, which offers a scientific basis for choosing the appropriate satellite imagery to create accurate LULC maps.

Translation Method of Emotional Expression in Daniel Keyes’ The Minds of Billy Milligan

This study examines the English to Indonesian translation of emotional expressions depicted in Daniel Keyes’s novel “The Minds of Billy Milligan” and its Indonesian translation “24 Wajah Billy.” The primary objectives are to describe the emotional expressions portrayed by the main character and to analyze the methods applied in translating these emotional expressions from English to Indonesian. Employing David Krech’s emotional classification theory and drawing on Vinay and Darbelnet’s translation methods, this qualitative descriptive research identifies 28 emotional classification data in the novel. Eight types of emotion classifications were identified, including 13 instances of fear, 6 instances of anger, 3 instances of hate, 2 instances of happiness, 1 instance of pain, 1 instance of guilt and regret, 1 instance of pride and shame, and 1 instance of love. The translation applies six translation methods used in 39 instances borrowing, literal, transposition, modulation, equivalent, and adaptation. Each emotion is translated using different translation methods. The application of these six methods is divided into three variants: single with 18 instances, couplet with 9 instances, and triplet with 1 instance. The literal method is the most frequently used in translating emotional expression. The translation method aims to preserve the essence of the narrative from the source language to the target language.