جامعة المعرفة
المصدر المعتمد للمعلومات البحثية والمخرجات هو قاعدة بيانات سكوبس (Elsevier Scopus) مع تدقيق بيانات الباحثين من عمادة البحث العلمي.
AD

Ashit Kumar Dutta

اشيت كومار دوتا

— Computer Science

كلية الهندسة وعلوم الحاسببانتظار توثيق الانتساب2021–2026
220وثائق Scopus
4,399استشهاداً
27مؤشر h
203منها بانتماء الجامعة
إجمالي Scopus (كل الجهات)ملف الباحث في Scopus ↗آخر تحديث: 2026-11-09

كل وثائق Scopus للباحث

220
مقالوصول مفتوح20248استشهاد· Semantic Scholar

A Speech Disorder Detection Model Using Ensemble Learning Approach

Journal of Disability Research · مجلد 3 (3)

Speech disorder detection (SDD) models can assist speech therapists in providing personalized treatment to individuals with speech impairment. Speech disorders (SDs) comprise a broad spectrum of problems that affect the production, articulation, fluency, and quality of speech. Prompt identification and timely intervention are essential for efficient control and therapy of speech problems. However, current diagnostic techniques often depend on subjective evaluations by speech-language pathologists and may encounter challenges in terms of accessibility, scalability, and consistency. The limited dataset and substantial computational power reduce the efforts for developing automated SDD models. However, recent technology developments enable researchers to determine key factors to classify voice samples. The voice sample classification can identify the severity of the SD. Ensemble learning (EL) facilitates the integration of multiple models’ predictions to generate an optimal outcome. Hence, an EL-based SDD model is introduced in this study. A mel-spectrogram (MS) generator is built to produce images using the voice samples. The authors suggested a feature engineering based on the ResNet 18 model for extracting crucial features from the MS. CatBoost and XGBoost models are employed to classify the features. The outcomes of these models are used to train the support vector machine model to make a final prediction. The VOice ICarfEDerico II (VOICED) dataset is used to generalize the proposed model. The experimental findings indicate that the recommended SDD model outperforms the state-of-the-art models by yielding an outstanding performance. This suggests that the model can assist speech therapists in offering customized speech therapies based on the SD severity. The model performance can be improved using multilanguage voice samples.

ensemble learningMel spectrogramCatBoostXGBoost
مقالوصول مفتوح20248استشهاد· Semantic Scholar

A Fine-Tuned CatBoost-Based Speech Disorder Detection Model

Journal of Disability Research · مجلد 3 (3)

The classification of speech disorders (SDs) is crucial for treating children with speech impairment (SI). An automated SD classification can assist speech therapists in rendering services to children with SI in rural areas. Automated techniques for detecting SDs provide objective assessments of speech attributes, including articulation, fluency, and prosody. Clinical examinations and quantitative assessments provide an in-depth understanding of the patient’s speaking abilities and limitations. Existing deep learning (DL) models for SD detection often lack generalization across diverse populations and speech variations, leading to suboptimal performance when applied to individuals with different linguistic backgrounds or dialects. This study introduces a DL-based model for classifying normal and abnormal speeches using voice samples. To overcome the overfitting and bias, the authors construct convolutional neural network models with the weights of MobileNet V3 and EfficientNet B7 models for feature extraction (FE). To improve performance, they integrate the squeeze and excitation block with the MobileNet V3-based FE model. Similarly, the EfficientNet B7-model-based FE is improved using the structure pruning technique. The enhanced CatBoost model differentiates the normal and abnormal speeches using the extracted features. The experimental analysis is performed using the public dataset that contains 4620 utterances of healthy children and 2178 utterances of children with SI. The comparative study reveals the exceptional performance of the proposed SD classification model. The model outperforms the current SD classification models. It can be employed in clinical settings to support speech therapists. Substantial training with diverse voice samples can improve the generalizability of the proposed model.

CatBoostconvolutional neural network feature extractionMobileNetV3EfficientNetB7
مقالوصول مفتوح20247استشهاد· Semantic Scholar

Feature Fusion-based Brain Stroke Identification Model Using Computed Tomography Images

Journal of Disability Research · مجلد 3 (5)

Accurate and rapid diagnosis is essential in the healthcare system for the detection of strokes to mitigate the devastating effects. This study introduces an innovative model for identifying strokes using advanced deep learning (DL) architectures, including SqueezeNet v1.1 and MobileNet V3-Small, feature fusion approaches, and CatBoost models. Using SqueezeNet v1.1 and MobileNet V3-Small, the authors extract meaningful features from computed tomography images that capture local details and global patterns suggesting stroke conditions. Subsequently, they employ feature fusion to combine the complementary representations derived by both architectures, consequently boosting the discriminative capability of the feature set. The Optuna-based CatBoost model is employed to predict stroke using the fused features. The experimental findings show outstanding performance, with a considerable accuracy of 99.1%. The high accuracy level demonstrates our suggested method’s effectiveness in precisely detecting strokes from medical imaging data. Combining DL architectures, feature fusion, and gradient-boosting models offers a promising approach to enhancing stroke diagnosis systems. This can potentially improve patient outcomes and clinical decision-making in stroke treatment.

stroke detectionCT imagesfeature fusionCatBoost
مقالوصول مفتوح20247استشهاد· Semantic Scholar

Thermal/econmic/environmental considerations in a multi-geneation layout with a heat recovery process; A multi-attitude optimization based on ANN approach

Case Studies in Thermal Engineering · مجلد 55 · 104170-104170

The proposed system is designed to address the growing demand for sustainable and efficient energy solutions that can minimize greenhouse gas emissions while meeting the increasing energy needs of a modern world. The integrated approach of the system allows it to produce not only electricity but also hydrogen and fresh water from seawater, making it a comprehensive solution to multiple challenges. The system comprises several components that work together including a gas turbine unit, Rankine and organic Rankine cycles, a multi-effect desalination (MED) system, and a proton exchange membrane (PEM) electrolyzer. One significant characteristic of this system is its ability to generate hydrogen, which is later combined with the inlet fuel of the gas turbine. This novel methodology presents a more environmentally friendly fuel alternative, showcasing its potential to enhance combustion efficiency in a sustainable manner. To ensure the viability and feasibility of the system, a MATLAB function was utilized for calculating the base case scenario, followed by a GA optimization technique with the aid of a machine learning model to minimize run-time. In the given scenario, the total cost rate is measured at 3726 $/h, with an exergy efficiency of 54% and a normalized CO2 emission of 305 kg/GJ. Following the optimization process, significant enhancements can be observed. The cost rate decreases by 2202 $/h, the exergy efficiency improves from 54% to 55.61%, and the annual CO2 emissions decrease from 133,760 metric tons to 50,320 metric tons. Moreover, the implementation of blended fuel leads to a 10% decrease in normalized CO2 emissions.

multi-generation systemheat recoverygas turbinesorganic Rankine cycle
مقالوصول مفتوح20247استشهاد· Semantic Scholar

Ensemble Learning-based Alzheimer’s Disease Diagnosis Using Magnetic Resonance Imaging

Journal of Disability Research · مجلد 3 (6)

The progressive nature and early identification requirements of Alzheimer’s disease (AD) provide an immense challenge in healthcare. The present study introduces a novel ensemble learning technique for detecting AD, using cutting-edge deep learning (DL) and classic machine learning (ML) techniques. The feature extraction process is carried out with YOLOv7 and EfficientNet B3 models, which effectively capture spatial and semantic information from brain imaging data. CatBoost and XGBoost are used as base learners, using gradient-boosting capabilities for classification. In order to improve the accuracy of predictions, support vector machines are used as meta-learners to effectively merge the results of the base models. We performed trials on a dataset from the Kaggle repository and achieved a remarkable average accuracy of 99.8%. Our methodology shows the effectiveness of integrating DL and classic ML methods in detecting AD. The ensemble architecture not only boosts the accuracy of classification but also improves the resilience and generalizability of the model. The study’s results indicate promising directions for advancing the development of precise and dependable diagnostic instruments for AD. The proposed research has the potential to assist medical professionals in identifying the condition at an early stage and planning appropriate interventions and treatments.

Alzheimer's disease diagnosisensemble learningbrain MRIYOLOv7
مقالوصول مفتوح20246استشهاد· Semantic Scholar

SADO-Net: A spatial adaptive dart optimized network model for an automated brain tumor diagnosis using MRIs

Alexandria Engineering Journal · مجلد 109 · 884-902

Early detection of brain tumors is essential, and a biopsy is required to determine the type of tumors in the brain and is only possible after extensive brain surgery. Brain tumors can be identified and classified by clinicians with the aid of computational intelligence algorithms. Here, we developed an novel and intelligent automated system, known as, Spatial Adaptive Dart Optimized Network (SADO-Net) for the diagnosis of brain tumor using MRIs. This will allow medical professionals to identify tumors in their early stages with a high degree of accuracy. Before classifying the diseases, the preprocessed brain MRIs are segmented using the Spatial Pattern based Image Segmentation (SPISeg) technique. The type of brain tumor is then promptly and reliably identified using the Weight Optimized Deep Network (WODNet) Classification model. The adoption of the Darts Game based Optimization (DGO) method for feature reduction expedites the classification process. This work uses a range of metrics and popular public datasets such as BRATS 2018, BRATS 2019, BRATS 2020 and Figshare to compare and validate the performance of the proposed SADO-Net model in tumor identification. Based on the results, the SADO-Net model exhibits good performance, with an average accuracy of 99.2 % and a loss rate of 0.5 %.

brain tumor classificationautomated diagnosismagnetic resonance imagingfeature reduction
مقالوصول مفتوح20246استشهاد· Semantic Scholar

Artificial intelligence probabilities scheme for disease prevention data set construction in intelligent smart healthcare scenario

SLAS TECHNOLOGY · مجلد 29 (4) · 100164-100164

In the face of an aging population, smart healthcare services are now within reach, thanks to the proliferation of high-speed internet and other forms of digital technology. Data problems in smart healthcare, unfortunately, put artificial intelligence in this area to serious limitations. There are several issues, including a lack of standard samples, noisy data interference, and actual data that is missing. A three-stage AI-based data generating strategy is suggested to handle missing datasets, using a small sample dataset obtained from a smart healthcare program community in a specific city: Step one involves generating the dataset's basic attributes using a tree-based generation strategy that takes the original data distribution into account. Step two involves using the Naive Bayes algorithm to create basic indicators of behavioural capability assessment for the samples. Step three builds on stage two and uses a multivariate linear regression method to create evaluation criteria and indicators of high-level behavioural capability. Six problems involving multiple classifications and two tasks using multiple labels are implemented using various neural network-based training strategies on the obtained data to assess the usefulness of the dataset for downstream tasks. To ensure that the data collected is genuine and useful, the experimental data must be analysed and expert knowledge must be included.

synthetic data generationmissing data imputationsmart healthcaredisease prevention
مقالوصول مغلق20245استشهاد· Semantic Scholar

Predictive Monitoring in Process Mining Using Deep Learning for Better Consumer Service

IEEE Transactions on Consumer Electronics · مجلد 70 (4) · 7279-7290

Process mining, a burgeoning discipline within data science, demonstrates a significant contribution to the software development lifecycle of diverse real-time consumer-centric projects. This paper underscores the prominence of integrating predictive business process monitoring into organizational process models, as it can substantially impact profits and efficiency in any possible business domain along with improving services to consumers. The paper proposes a novel deep learning-based business process prediction model consisting of multiple layers with fine-tuning hyperparameters. The proposed model leverages input embeddings to represent each of the activities, and based on the training of the proposed model, the accuracy of the next activity is calculated. To assess the efficacy of the proposed model, it has been compared with the existing benchmark models. Our proposed model has shown a significant gain over the existing approaches. The results show that the proposed model outperforms these approaches by achieving an accuracy of 76% on the consumer helpdesk dataset along with an accuracy of 78% on the benchmark BPI dataset.

process miningpredictive monitoringnext activity predictiondeep learning
مقالمفتوح (مستودع)20245استشهاد· Semantic Scholar

Optimizing hybrid deep learning models for drug‐target interaction prediction: A comparative analysis of evolutionary algorithms

Expert Systems · مجلد 42 (2)

Abstract In the realm of Drug‐Target Interaction (DTI) prediction, this research investigates and contrasts the efficacy of diverse evolutionary algorithms in fine‐tuning a sophisticated hybrid deep learning model. Recognizing the critical role of DTI in drug discovery and repositioning, we tackle the challenges of binary classification by reframing the problem as a regression task. Our focus lies on the Convolution Self‐Attention Network with Attention‐based bidirectional Long Short‐Term Memory Network (CSAN‐BiLSTM‐Att), a hybrid model combining convolutional neural network (CNN) blocks, self‐attention mechanisms, and bidirectional LSTM layers. To optimize this complex model, we employ Differential Evolution (DE), Particle Swarm Optimization (PSO), Memetic Particle Swarm Optimization Algorithm (MPSOA), Fire Hawk Optimization (FHO), and Artificial Hummingbird Algorithm (AHA). Through thorough comparative analysis, we evaluate the performance of these evolutionary algorithms in enhancing the CSAN‐BiLSTM‐Att model's effectiveness. By examining the strengths and weaknesses of each algorithm, our study aims to provide valuable insights into DTI prediction, identifying the most effective evolutionary algorithm for hyperparameter tuning in advanced deep learning models. Notably, Fire‐hawk optimization (FHO) emerges as particularly promising, achieving the highest Concordance Index (C‐index) as 0.974 for KIBA datasets and 0.894 for DAVIS datasets and demonstrating exceptional accuracy in ranking continuous predictions across both the datasets.

drug-target interaction predictionevolutionary algorithmshyperparameter optimizationhybrid deep learning model
مقالوصول مفتوح20245استشهاد· Semantic Scholar

Developing a Pain Identification Model Using a Deep Learning Technique

Journal of Disability Research · مجلد 3 (3)

In this study, the authors proposed a pain identification model using facial expressions. An image extraction technique was developed using the liquid neural network to extract diverse images from the video files. The authors used the DenseNet 201 and MobileNet V3 models to build a hybrid feature engineering technique. They applied quantization aware training to improve the efficiency of the models. The Prkachin and Solomon Pain Intensity score was used for the image classification. They fine-tuned the LightGBM model using the random search algorithm for identifying pain from the facial images. The authors used the Denver Intensity of Spontaneous Facial Action dataset to generalize the proposed model. The performance evaluation outlined the significant performance of the proposed model in identifying pain using the images. In addition, it demands limited computational resources to identify pain. Healthcare and rehabilitation centers can implement the proposed model to provide adequate services to disabled individuals.

pain intensity estimationfacial expression analysisfacial action unitsliquid neural networks