Preprocessing Step Research Articles

Research on multimodal English teaching model based on genetic algorithm

In recent years, English language teaching has undergone significant transformation due to the integration of technology and innovative teaching methodologies. Among the most promising advancements is the formation of multimodal teaching, which incorporates various forms of media. Despite the potential benefits associated with multimodal teaching, challenges remain in tailoring content to individual learning needs and optimizing the learning process for maximum effectiveness. To address these challenges, this research proposes a Multimodal English Teaching Framework (METF) to enhance the effectiveness and personalization of language learning. The framework integrates multiple teaching modalities, including text, audio, video, and interactive elements, to accommodate diverse learning preferences. It incorporates student interaction data (responses, time spent on tasks, and engagement levels), learner demographics, learning styles, and feedback records, thereby capturing the effectiveness of personalized content delivery. Pre-processing steps, including tokenization for text data, noise reduction for audio data, and normalization for resolution and interaction data, ensure data quality and consistency. Feature extraction techniques such as Convolutional Neural Networks (CNNs) for audio and video data and Term Frequency-Inverse Document Frequency (TF-IDF) for text data are applied to capture meaningful patterns. To enhance the fusion of multimodal data, feature-level fusion is implemented, integrating recovered characteristics from multiple modalities into a unified representation that supports more accurate decision-making. This research establishes the Resilient Genetic Algorithm (RGA) to enhance the adaptive learning process by ensuring robustness and avoiding premature convergence. Experimental results demonstrate the potential of this multimodal approach, achieving 94.2% accuracy, a convergence time of 45%, task engagement of 89.4%, and personalization effectiveness of 91.7%. These results indicate a high value of accuracy in multimodal English teaching. This research highlights the applicability of algorithms in optimizing complex, multimodal educational environments.

Early Parkinson’s disease identification via hybrid feature selection from multi-feature subsets and optimized CatBoost with SMOTE

Achieving high accuracy, efficiency and robustness remains a primary challenge in Parkinson's disease (PD) detection, as existing methods often struggle with these aspects. Additionally, data imbalance in medical datasets further limits the reliability of current models. Given the critical role of precise PD classification in medical diagnostics, this study proposes a novel framework to enhance detection accuracy. The proposed framework leverages a strong categorical boosting (CatBoost) algorithm optimized using Grid Search Optimization (GSO). The analysis was conducted on a PD dataset derived from speech recording signals. To address the data imbalance, the synthetic minority oversampling technique (SMOTE) is applied as a pre-processing step to improve the robustness and reliability of the model. The framework is tested with several feature subsets, as well as their combined set. In addition, RReliefF feature selection was also applied to identify the optimal subset of features. For the selected feature subset, the proposed approach achieved an accuracy of 0.9261, precision of 0.9633, sensitivity of 0.9375, F1-score of 0.9502, specificity of 0.8947, AUC of 0.9549 and a testing time of 0.012s. These results underscore the effectiveness of the proposed framework in accurately diagnosing PD, thereby providing a reliable and efficient tool for future medical diagnostics.

LED array position error correction in lensless optical microscopy through adaptive genetic strategy

Abstract Despite the complex optical lens system, classical optical microscopy cannot achieve high spatial resolution and a large field of view (FOV). The lensless optical ptychographic microscopy technique solves the contradiction between resolution and FOV with a simple optical path. However, position errors between LED array and sample, caused by misalignment of optical imaging axis and manufacture errors of LED array, decrease reconstruction quality. Due to the difficulty of hardware correction, we propose an adaptive genetic strategy to correct LED array position error for lensless ptychographic imaging. Firstly, the coarse positions of the LEDs are predicted by Kalman filter, thus we can determine the position range. Secondly, the precise positions of LEDs are calculated by our adaptive genetic algorithm, which combines the fitness function according to the previous predicted positions. Finally, we incorporate the LED array position correction into the Gerchberg–Saxton (GS) phase recovery method to achieve improved reconstructed image quality. Compared with similar methods, both simulation and real experiments indicate that our method genetic algorithm adapt-GS (adaptive Genetic algorithm-based GS) method has higher robustness in position errors correction, and creates results with better quality. This method is applied to other state-of-art reconstruction algorithms as a pre-processing step, highly improve the image quality.

A Hybrid Feature Selection Approach on Medical Dataset

Feature Selection and outlier detection are one of the most crucial steps in machine learning, where the priority is to simultaneously improve classification accuracy and decrease computation time by selecting the most relevant features of the dataset and removing those data instances which behave abnormally from the other data entities without sacrificing the critical information carried out by the data. This research employs a hybrid ensemble filter and wrapper strategy to select significant features; before that, as a pre-processing step to increase the accuracy of the classifier, a cluster-based outlier elimination approach is employed from 15 benchmark medical datasets including COVID-19 dataset. In the medical and healthcare sectors, the advent of modern technology has led to an increase in the volume of diverse data kinds, from which some of the features and instances are irrelevant and noisy, which significantly degrades the classifier’s performance and hence leads to a poor decision made over these vital datasets. Therefore, feature selection becomes inevitable. Compared to a previously described wrapper technique and four distinct filter methods for accuracy level, precision, recall, F1-measure and size of the chosen feature subset, the proposed approach yields better results.

Survival prediction from imbalanced colorectal cancer dataset using hybrid sampling methods and tree-based classifiers

Colorectal cancer is a high mortality cancer, with a mortality rate of 64.5% for all stages combined. Clinical data analysis plays a crucial role in predicting the survival of colorectal cancer patients, enabling clinicians to make informed treatment decisions. However, utilizing clinical data can be challenging, especially when dealing with imbalanced outcomes, an aspect often overlooked in this context. This paper focuses on developing algorithms to predict 1-, 3-, and 5-year survival of colorectal cancer patients using clinical datasets, with particular emphasis on the highly imbalanced 1-year survival prediction task. We utilized a colorectal cancer dataset from the Surveillance, Epidemiology, and End Results (SEER) database, which exhibits high imbalance in the 1-year (1:10) survival analysis and an imbalance in the 3-year (2:10) analysis, achieving balance in the 5-year analysis. The pre-processing step consists of removing records with missing values and merging categories with less than 2% share for each categorical feature to limit the number of classes of each component. Edited Nearest Neighbor, Repeated Edited Nearest Neighbor (RENN), Synthetic Minority Over-sampling Technique (SMOTE), and pipelines of SMOTE and RENN approaches were used for balancing the data with tree-based classifiers, including Decision Tree, Random Forest, Extra Tree, eXtreme Gradient Boosting, and Light Gradient Boosting Machine (LGBM). The performance evaluation utilizes a 5-fold cross-validation approach. In the case of 1-year, our proposed method with LGBM significantly outperforms other sampling methods with the sensitivity of 72.30%. For the task of 3-year survival, the combination of RENN and LGBM achieves a sensitivity of 80.81%, indicating that our proposed method works best for highly imbalanced datasets. Additionally, when predicting 5-year survival, the sensitivity reaches 63.03% using LGBM. Our proposed method significantly improves mortality prediction for the minority class of colorectal cancer patients. RENN followed by SMOTE yields better sensitivity in the classifiers, with LGBM as the predictor performing best for 1- and 3-year survival. In the 5-year task, LGBM outperforms other models in terms of F1-score.

Advanced Software Techniques for Detecting Digital Image Manipulation

It is necessary to detect forgery of digital images in order to maintain their integrity. In this paper an attempt is made to solve the problem of copy-move forgery detection which is the most common form of image manipulation. We propose two new methods for detecting duplicated regions which are based on the texture and statistical features. The first technique is based on the classical SIFT (Scale-Invariant Feature Transform) which is a keypoint based method; the second one is based on the integration of SIFT into deep learning techniques and software solutions. We used the MICC-F600 database to evaluate the proposed methods and split it into training, validation and test sets. To increase performance of the model, some pre-processing steps were included such as scaling, image polishing, and so on. Based on the experimental results, our software-embedded convolutional neural network (CNN) models reached the greatest accuracy in identifying forged images. The model develops a binary mask that estimates the location of forgery in an image, which assists to improve the precision of digital forensic examination with our intelligent software.

Deep learning models for segmenting phonocardiogram signals: a comparative study.

Cardiac auscultation requires the mechanical vibrations occurring on the body's surface, which carries a range of sound frequencies. These sounds are generated by the movement and pulsation of different cardiac structures as they facilitate blood circulation. Subsequently, these sounds are identified as phonocardiogram (PCG). In this research, deep learning models, namely gated recurrent neural Network (GRU), Bidirectional-GRU, and Bi-directional long-term memory (BILSTM) are applied separately to segment four specific regions within the PCG signal, namely S1 (lub sound), the systolic region, S2 (dub sound), and the diastolic region. These models are applied to three well-known datasets: PhysioNet/Computing in Cardiology Challenge 2016, Massachusetts Institute of Technology (MITHSDB), and CirCor DigiScope Phonocardiogram.The PCG signal underwent a series of pre-processing steps, including digital filtering and empirical mode decomposition, after then deep learning algorithms were applied to achieve the highest level of segmentation accuracy. Remarkably, the proposed approach achieved an accuracy of 97.2% for the PhysioNet dataset and 96.98% for the MITHSDB dataset. Notably, this paper represents the first investigation into the segmentation process of the CirCor DigiScop dataset, achieving an accuracy of 92.5%. This study compared the performance of various deep learning models using the aforementioned datasets, demonstrating its efficiency, accuracy, and reliability as a software tool in healthcare settings.

Interval Boundary Correction for Interval Linear Program

An interval linear programming problem is a bounded problem when all of its deterministic problems are bounded. However, if the best deterministic problem is unbounded or the worst deterministic problem is infeasible, a pre-processing step on tightening interval parameter boundaries should be employed to obtain an adjusted bounded interval linear program. With the help of the duality theorem, we propose an interval boundary correction algorithm designed to minimize the loss of interval information. The algorithm makes relatively small adjustments with respect to the interval width of each interval parameter boundary. Some numerical examples illustrating this correction algorithm are provided.

Decoding diffraction and spectroscopy data with machine learning: A tutorial

This Tutorial provides a step-by-step guide on how to apply supervised machine-learning techniques to analyze diffraction and spectroscopy data. This Tutorial details four models—a reconstruction-focused model, a regression-focused model, a hybrid reconstruction/regression model, and a multimodal model—that use x-ray diffraction profiles and vibrational density of states spectra to predict various microstructural descriptors. In this Tutorial, we cover data pre-processing steps, constructions of the models via dimensionality reduction and regression, training, and analysis of these models. Comparisons of the model’s performance are provided, highlighting the strength and weakness of the various approaches utilized.

Estimating sensor-space EEG connectivity PART 2: Identifying optimal artifact reduction techniques for functional connectivity in real data.

Estimating sensor-space EEG connectivity PART 2: Identifying optimal artifact reduction techniques for functional connectivity in real data.

Irish potato imagery dataset for detection of early and late blight diseases.

Irish potato imagery dataset for detection of early and late blight diseases.

Natural language processing techniques applied to the electronic health record in clinical research and practice - an introduction to methodologies.

Natural Language Processing (NLP) has the potential to revolutionise clinical research utilising Electronic Health Records (EHR) through the automated analysis of unstructured free text. Despite this potential, relatively few applications have entered real-world clinical practice. This paper aims to introduce the whole pipeline of NLP methodologies for EHR analysis to the clinical researcher, with case studies to demonstrate the application of these methods in the existing literature. Essential pre-processing steps are introduced, followed by the two major classes of analytical frameworks: statistical methods and Artificial Neural Networks (ANNs). Case studies which apply statistical and ANN-based methods are then provided and discussed, illustrating information extraction tasks for objective and subjective information, and classification/prediction tasks using supervised and unsupervised approaches. State-of-the-art large language models and future directions for research are then discussed. This educational article aims to bridge the gap between the clinical researcher and the NLP expert, providing clinicians with a background understanding of the NLP techniques relevant to EHR analysis, allowing engagement with this rapidly evolving area of research, which is likely to have a major impact on clinical practice in coming years.

Forecasting Dissolved Oxygen in Lakes Using Different AI Models

The freshwater ecological situation and water quality management would be impossible without the accurate prediction of the dissolved oxygen (DO) concentration in lakes. This study focuses on the use of different advanced Artificial Intelligence (AI) models in predicting values of the DO level for a given input environmental data. To make the data usable for analysis, a number of pre-processing steps are carried out. These processes include, but are not limited to, the ability to deal with the missing data and standardization of the features in the data set, so that the features are on the same level. After the pre-processing, the dataset aggregated in features and a target variable, which is the concentration of the target substance, the dissolved oxygen, is selected. Also, as a part of the building process of the model, the analysis of the features with regard to the target variable is performed in order to discern contributing features for better prediction. Additionally, in order to increase the accuracy of the machine learning models, standardization transformations to the mean of zero and to the variance of one were applied in the training set and test set using Standard Scaler. A selection of different machine learning models was therefore performed in order to identify the best predictor of the DO concentration. In the case of Linear Regression, it’s R² has indicated a high degree of predictive accuracy at 0.9974 with a very low Mean Squared Error (MSE) of 9.6146e-05. Meagerly performing as an alternate method, Support Vector Regression (SVR) managed to attain 0.9448 of R² and an MSE of 0.0021, so it works but not as accurately as Linear Regression. In addition, multiple hidden layers were applied through Artificial Neural Networks (ANN) in a bid to model the data as uncertainty that appears in higher orders. Most notably, the ANN model achieved an R² of 0.995 and an MSE of 0.000161 which is nearly comparable to that of Linear Regression.

Tribological Performance of AlCrN and TiAlN Coatings in Aluminum Forming Process at High Temperature

Material transfer is a critical tribological issue in aluminum forming processes, particularly at high temperatures. Although applying PVD or CVD coatings on the forming tool surface could be a solution to alleviate the material transfer issue, selecting an appropriate coating with sufficient tribological performance is still a challenging task. This paper aims to evaluate the tribological performance of AlCrN and TiAlN coatings against 6082-T6 aluminum alloy under dry contact conditions at 400°C by using a process centered approach. The methodological approach implies: 1. pre-processing to assess the contact conditions to be simulated and prepare the experimental samples, 2. processing where the friction tests are performed, and 3. post-processing in which test results are analyzed in terms of coating performance. The warm and hot upsetting sliding test (WHUST) is used as a tribometer in the processing step. A scaled-down version of the existing WHUST is developed to integrate into the heating chamber of Bruker UMT TriboLab for precise control of the specimen and contactor temperatures. Numerical simulation is applied to identify the WHUST parameters in the pre-processing step to reproduce severe contact conditions from the industrial forming processes. The post-processing step includes the identification of Coulomb’s coefficient of friction (COF) and the surface analysis with quantitative and qualitative techniques.

A novel method for brain aneurysms computed fluid dynamics analysis after flow diverter stent implantation based on micro-computed tomography reconstruction.

Flow diverter stents are now increasingly utilised in the management of cerebral aneurysms. Computed flow simulation can be used to assess flow reduction and anticipate treatment outcome. To achieve this, the stent must be deployed virtually in the parent vessel carrying the aneurysm before running the simulation. This study proposes an alternative method to virtual deployment using micro-computer tomography after in vitro deployment of stents in silicone models to reconstruct accurately the struts over the aneurysm neck. Two experienced neuroradiologists deployed different stents (four 48-wire and one 64-wire pipeline embolization devices) in identical silicone models. Micro-CT acquisition was then carried out and allowed to reconstruct the stent meshes very accurately using different pre-processing steps. Virtual deployments were also performed for the two types of stents. Qualitatively, it was shown that the real deployments formed an inhomogeneous wire network compared to the regularity of the mesh of the virtually deployed stents. This difference had a moderate influence on hemodynamics. For virtual deployment, we observed an overestimation of the flow reduction of 5% and 16% for the 64- and 48-wire stents, respectively. While the flow patterns remained the same for the 48-wire stent, the 64-wire stent showed an inversion of the intra-aneurysmal vortex. This promising method could be used in the development of flow diverter devices or to conduct a comparative evaluation between different devices.

A 2-level cascaded deep learning with the adaptive novel feature extraction approach for sales forecasting framework

This study presents a revolutionary deep-learning architecture that focuses on feature extraction, feature selection, and sales forecasting. The technique begins with a pre-processing step using median imputation and min–max. To extract features, the study uses the Word2Vec (CBOW) model to construct contextual word embeddings and TF-IDF to assess term relevance in texts. For numerical data, the Pearson Correlation Coefficient (PCC) detects linear associations, whereas the Adaptive Bhattacharyya Correlation refines characteristics using statistical metrics. The Tent Chaotic-based Puma Optimizer Algorithm (TC-POA) is used to pick important characteristics for the model effectively. The forecasting challenge employs the 2-level Cascaded Deep-Learning (2-LCDL) model. In this model, an Alexnet with Squeeze-and-Excitation is used along with the Bi-LSTM to capture sequential data dependencies and GRU for efficient long-term memory management. The work wraps up with a fully linked layer and SoftMax activation for sales prediction. This framework demonstrates substantial advances in deep-learning approaches, providing better-predicted accuracy and computational efficiency. The proposed model is evaluated for both 70% and 80% of training data for the metrics MSE of 0.0058, 0.0054, MAPE of 1.1968, 0.9979 and also the RMSE, G-Mean, MAE, MMAE, MNB, IoA have superior performance when compared to the other existing methods.

Analysis of the BraTS 2023 Intracranial Meningioma Segmentation Challenge

We describe the design and results from the BraTS 2023 Intracranial Meningioma Segmentation Challenge. The BraTS Meningioma Challenge differed from prior BraTS Glioma challenges in that it focused on meningiomas, which are typically benign extra-axial tumors with diverse radiologic and anatomical presentation and a propensity for multiplicity. Nine participating teams each developed deep-learning automated segmentation models using image data from the largest multi-institutional systematically expert annotated multilabel multi-sequence meningioma MRI dataset to date, which included 1000 training set cases, 141 validation set cases, and 283 hidden test set cases. Each case included T2, FLAIR, T1, and T1Gd brain MRI sequences with associated tumor compartment labels delineating enhancing tumor, non-enhancing tumor, and surrounding non-enhancing FLAIR hyperintensity. Participant automated segmentation models were evaluated and ranked based on a scoring system evaluating lesion-wise metrics including dice similarity coefficient (DSC) and 95% Hausdorff Distance. The top ranked team had a lesion-wise median dice similarity coefficient (DSC) of 0.976, 0.976, and 0.964 for enhancing tumor, tumor core, and whole tumor, respectively and a corresponding average DSC of 0.899, 0.904, and 0.871, respectively. These results serve as state-of-the-art benchmarks for future pre-operative meningioma automated segmentation algorithms. Additionally, we found that 1286 of 1424 cases (90.3%) had at least 1 compartment voxel abutting the edge of the skull-stripped image edge, which requires further investigation into optimal pre-processing face anonymization steps.

ViTBI-BERT: A Vietnamese Language Model for Prediction of Traumatic Brain Injury

Artificial Intelligence (AI) plays an increasingly significant role in predicting diseases by leveraging large-scale and complex datasets. AI-based traumatic brain injury (TBI) prediction has traditionally relied on multimodal data, including both structured information like test results and unstructured data such as CT or/and MRI scans. However, one important source of information—physician conclusions in text form at the time of admission and discharge—has largely been underutilized. These textual conclusions offer valuable insights into the patient's condition. This paper proposes a novel approach to classifying TBI severity using physicians' written conclusions. We introduce the viTBI-BERT model, which is built on the ViHealthBERT backbone. Our approach includes pre-processing techniques such as acronym normalization, special character removal, and noun chunking and textual data augmentation. After pre-processing step, the data is passed through the pre-trained ViHealthBERT model, with an additional fully connected layer and a softmax layer for the classification task. The model is tested on three sets of medical notes derived from a self-collected dataset consisting of clinical and CT findings from 503 Vietnamese patients. The viTBI-BERT model achieved a highest sensitivity of 71% for classifying four levels of injury severity. These findings demonstrate the potential of language-based analysis, which, when combined with structured clinical and subclinical data, could further enhance TBI classification outcomes.

Evaluating the Effectiveness of CNN, LSTM, and Bi-LSTM Models in Classifying Twitter Sentiments

Introduction: Twitter sentiment analysis is an essential tool for understanding public opinions and extracting valuable insights from social media discussions. However, the informal, concise nature of tweets, along with their context-dependent language, poses significant challenges for accurate sentiment classification. Deep learning techniques have shown promise in addressing these challenges due to their ability to learn complex patterns and contextual relationships in data. This study explores the application of Convolutional Neural Networks (1D-CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (Bi-LSTM) models for classifying tweet sentiments into four categories: positive, negative, neutral, and irrelevant. Objectives: The primary objective of this study is to evaluate the effectiveness of deep learning models, including 1D-CNN, LSTM, and Bi-LSTM, in classifying tweet sentiments into categories such as positive, negative, neutral, and irrelevant. By analyzing the performance of these models, the study aims to compare their accuracy, precision, recall, and F1-score to determine their strengths and limitations. Additionally, the research seeks to identify the most suitable model capable of effectively capturing the sequential and contextual nature of tweets, addressing the unique challenges posed by the informal and context-dependent language of social media data. Methods: The study employs a publicly available Twitter sentiment dataset comprising 73,906 tweets related to general Twitter discussions. The dataset undergoes pre-processing steps, including tokenization, stopword removal, and handling emoticons and hashtags, to ensure clean input for training the models. Tweets are represented using pre-trained word embeddings, specifically GloVe and Word2Vec, which provide rich semantic information by capturing word meanings in context. Individual models—1D-CNN, LSTM, and Bi-LSTM—are trained and evaluated using this pre-processed data. Performance metrics, including accuracy, precision, recall, and F1-score, are calculated to compare the models' effectiveness. Results: The experimental analysis demonstrates that each model has unique strengths in sentiment classification. However, the 1D-CNN outperforms both LSTM and Bi-LSTM models, achieving superior results in capturing both the sequential and contextual information inherent in tweet data. This highlights its efficiency and suitability for Twitter sentiment analysis. Conclusions: This study underscores the potential of deep learning techniques for sentiment analysis in social media. Among the evaluated models, 1D-CNN proves to be the most effective, offering a robust approach to handling the complexities of tweet sentiment classification. Future work could explore hybrid models and further optimization techniques to enhance sentiment analysis performance.

Proving as Fast as Computing: Succinct Arguments with Constant Prover Overhead

Succinct arguments are proof systems that allow a powerful, but untrusted, prover to convince a weak verifier that an input x belongs to a language \(L \in \mathsf {NP} \) , with communication that is much shorter than the \(\mathsf {NP} \) witness. Such arguments, which grew out of the theory literature, are now drawing immense interest also in practice, where a key bottleneck that has arisen is the high computational cost of proving correctness. In this work we address this problem by constructing succinct arguments for general computations, expressed as Boolean circuits (of bounded fan-in), with a strictly linear size prover. The soundness error of the protocol is an arbitrarily small constant. Prior to this work, succinct arguments were known with a quasi- linear size prover for general Boolean circuits or with linear-size only for arithmetic circuits, defined over large finite fields. In more detail, for every Boolean circuit C = C ( x , w ), we construct an O (log | C |)-round argument-system in which the prover can be implemented by a size O (| C |) Boolean circuit (given as input both the instance x and the witness w ), with arbitrarily small constant soundness error and using poly( λ , log | C |) communication, where λ denotes the security parameter. The verifier can be implemented by a size O (| x |) + poly( λ , log | C |) circuit following a size O (| C |) private pre-processing step, or, alternatively, by using a purely public-coin protocol (with no pre-processing) with a size O (| C |) verifier. The protocol can be made zero-knowledge using standard techniques (and with similar parameters). The soundness of our protocol is computational and relies on the existence of collision resistant hash functions that can be computed by linear-size circuits, such as those proposed by Applebaum et al. (ITCS, 2017). At the heart of our construction is a new information-theoretic interactive oracle proof ( \(\mathsf {IOP} \) ), an interactive analog of a \(\mathsf {PCP} \) , for circuit satisfiability, with constant prover overhead. The improved efficiency of our \(\mathsf {IOP} \) is obtained by bypassing a barrier faced by prior \(\mathsf {IOP} \) constructions, which needed to (either explicitly or implicitly) encode the entire computation using a multiplication code.