Manzoor Hussain
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Manzoor Hussain has a total of 57 co-authors. Here are the top co-authors by number of shared publications:
- Arun Malik
School of Computer Science and Engineering, Lovely Professional University, Phagwara 144411, India
3 shared publications - Abdul Salam Shah
Taylor's University, Malaysia
2 shared publications - Husnain Ali Haider
Department of Software Engineering, University of Sialkot, Sialkot 51040, Pakistan
2 shared publications - Isha Batra
Lovely Professional University
2 shared publications - Mehmood Naqvi
Mohwak College
2 shared publications - Rajeev Sobti
Lovely Professional University
2 shared publications
Latest Publications
- 16 October 2026
- Published by Springer Nature
- p. 436-447
Predicting Banana Quality Using Machine Learning Techniques
- 16 October 2026
- Published by Springer Nature
- p. 436-447
Predicting Banana Quality Using Machine Learning Techniques
- Open Access
- 27 May 2026
- Published by MDPI AG in Preprints.org
Analysis of Mild Overfitting, Class-Level Performance, and Future Directions in AI-Driven Fashion
The given research paper describes a CNN model of classifying images belonging to more than two classes on the Fashion-MNIST data. The model performed a test accuracy of 92.44% and test loss of 0.2533 the greatest accuracy as compared to similar studies with similar architectures. The architecture has three convolutional-pooling blocks, a dense layer with dropout regularization (0.3), and a softmax output layer. The analysis of training and validation curves demonstrates mild overfitting of the later epochs, and the validation loss starts growing even though the training loss continues to decrease. In-depth analysis using confusion matrix and classification report identifies certain patterns of misclassification between visually similar categories. The paper also discusses implications on batch normalization, data augmentation as well as Vision Transformer architecture.
- 27 May 2026
- Published by MDPI AG in Preprints.org
Open AccessAnalysis of Mild Overfitting, Class-Level Performance, and Future Directions in AI-Driven Fashion
The given research paper describes a CNN model of classifying images belonging to more than two classes on the Fashion-MNIST data. The model performed a test accuracy of 92.44% and test loss of 0.2533 the greatest accuracy as compared to similar studies with similar architectures. The architecture has three convolutional-pooling blocks, a dense layer with dropout regularization (0.3), and a softmax output layer. The analysis of training and validation curves demonstrates mild overfitting of the later epochs, and the validation loss starts growing even though the training loss continues to decrease. In-depth analysis using confusion matrix and classification report identifies certain patterns of misclassification between visually similar categories. The paper also discusses implications on batch normalization, data augmentation as well as Vision Transformer architecture.
- Open Access
- 27 May 2026
- Published by MDPI AG in Preprints.org
CNN-Based Fashion-MNIST Classification with Insights Toward AI-Based Mental Health Detection
The paper describes a Convolutional Neural Network which was trained on Fashion-MNIST with an accuracy of 89.57 (F1=0.90) on test with a three-block architecture of 32, 64 and 64 convolutional filters, using BatchNormalization and Dropout regularization, trained in 15 epochs using the Adam optimizer. The results of the experiment are rigorously applied to the field of AI-based mental health detection - among the most important and most actively evolving uses of deep learning in the health care industry around the world. More than one billion individuals have mental health disorders among them (WHO, 2022), and only the COVID-19 pandemic caused a 28% (193 to 246 million) and 25% (298 to 374 million) increase in depression and anxiety, respectively (PMC, 2024). CNNs are being actively used to detect mental health in three directions: facial expression recognition at depression and anxiety classification ( CNN+MDNet+ViT ensemble, PMC, 2024); speech spectral analysis at 93.5% depression severity grading and 91.2% depression episodes (PMC, 2024); and social media text classification with AI at 93.5% suicidal ideation and The CNN classification abilities verified on Fashion-MNIST include spatial feature hierarchy extraction, multi-class discrimination, softmax confidence estimation, and comprise the visual grounding layer of all three pathways. Privacy-preserving federated mental health AI, multimodal fusion, culturally adaptive models, longitudinal monitoring, and regulatory compliance are five of the future areas of research under the EU AI Act (2024).
- 27 May 2026
- Published by MDPI AG in Preprints.org
Open AccessCNN-Based Fashion-MNIST Classification with Insights Toward AI-Based Mental Health Detection
The paper describes a Convolutional Neural Network which was trained on Fashion-MNIST with an accuracy of 89.57 (F1=0.90) on test with a three-block architecture of 32, 64 and 64 convolutional filters, using BatchNormalization and Dropout regularization, trained in 15 epochs using the Adam optimizer. The results of the experiment are rigorously applied to the field of AI-based mental health detection - among the most important and most actively evolving uses of deep learning in the health care industry around the world. More than one billion individuals have mental health disorders among them (WHO, 2022), and only the COVID-19 pandemic caused a 28% (193 to 246 million) and 25% (298 to 374 million) increase in depression and anxiety, respectively (PMC, 2024). CNNs are being actively used to detect mental health in three directions: facial expression recognition at depression and anxiety classification ( CNN+MDNet+ViT ensemble, PMC, 2024); speech spectral analysis at 93.5% depression severity grading and 91.2% depression episodes (PMC, 2024); and social media text classification with AI at 93.5% suicidal ideation and The CNN classification abilities verified on Fashion-MNIST include spatial feature hierarchy extraction, multi-class discrimination, softmax confidence estimation, and comprise the visual grounding layer of all three pathways. Privacy-preserving federated mental health AI, multimodal fusion, culturally adaptive models, longitudinal monitoring, and regulatory compliance are five of the future areas of research under the EU AI Act (2024).
- Open Access
- 31 December 2025
- Published by VFAST Research Platform in VFAST Transactions on Software Engineering
- Vol. 13 (4) , 149-158
Enhancing Traffic Sign Recognition in Rough Weather to Reduce Traffic Accidents
Background: Automated traffic sign recognition is a technology that helps in detecting and understanding the traffic signs on the road. These signs may include stop signs, speed limits and some other warning signs. These signs are important in maintaining road safety and helping drivers obeying the traffic rules. Machine learning and computer vision together have advanced in this field and many methods have been proposed for automated traffic sign detection and recognition. Despite these advancements, there are still many challenging situations including occlusion, varying lighting conditions, difficult environmental conditions and sign variations which still need attention. Methods: We use Arabic traffic sign dataset to train EfficientNetB3 architecture with attention mechanism to classify the traffic sign under the unique visual and linguistic complexities as well as diverse environmental conditions. We also improved the dataset by augmenting and adding extra images to cover actual scenarios like fog, heavy rain, low light etc. making it stronger for testing and future research. Results: Our trained model achieved the accuracy of 99.61\% in testing better than the compared to the baseline CNN-Resnet model. Conclusion: In this research we addressed the existing limitations and sets a benchmark for the effective and efficient classification of Arabic traffic signs, particularly in challenging weather conditions.
- 31 December 2025
- Published by VFAST Research Platform in VFAST Transactions on Software Engineering
- Vol. 13 (4) , 149-158
Open AccessEnhancing Traffic Sign Recognition in Rough Weather to Reduce Traffic Accidents
Background: Automated traffic sign recognition is a technology that helps in detecting and understanding the traffic signs on the road. These signs may include stop signs, speed limits and some other warning signs. These signs are important in maintaining road safety and helping drivers obeying the traffic rules. Machine learning and computer vision together have advanced in this field and many methods have been proposed for automated traffic sign detection and recognition. Despite these advancements, there are still many challenging situations including occlusion, varying lighting conditions, difficult environmental conditions and sign variations which still need attention. Methods: We use Arabic traffic sign dataset to train EfficientNetB3 architecture with attention mechanism to classify the traffic sign under the unique visual and linguistic complexities as well as diverse environmental conditions. We also improved the dataset by augmenting and adding extra images to cover actual scenarios like fog, heavy rain, low light etc. making it stronger for testing and future research. Results: Our trained model achieved the accuracy of 99.61\% in testing better than the compared to the baseline CNN-Resnet model. Conclusion: In this research we addressed the existing limitations and sets a benchmark for the effective and efficient classification of Arabic traffic signs, particularly in challenging weather conditions.
- Open Access
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Machine Learning-Based Framework for Road Accident Detection and Prevention
As traffic density increases and road safety becomes more critical, there is an urgent need to develop predictive models and real-time systems for accident detection and prevention. A comprehensive dataset that includes factors such as vehicle speed, weather conditions, road type, and traffic flow is utilized in this research to analyze traffic patterns, environmental factors, and driver behavior in order to predict and prevent road accidents. These accidents cause loss of life, injuries, and economic damage. Early identification of accident conditions can mitigate risks, reduce response time, and save lives. We implemented supervised machine learning algorithms. These models were evaluated based on key performance metrics. Experiments were conducted using tools like RapidMiner and Python, and the Random Forest classifier achieved the highest accuracy of 94.8%. This model demonstrates superior predictive capabilities and robust performance in accident detection scenarios. The insights gained from this research contribute to developing intelligent systems for accident prevention, ensuring safer roads and reduced fatalities.
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessMachine Learning-Based Framework for Road Accident Detection and Prevention
As traffic density increases and road safety becomes more critical, there is an urgent need to develop predictive models and real-time systems for accident detection and prevention. A comprehensive dataset that includes factors such as vehicle speed, weather conditions, road type, and traffic flow is utilized in this research to analyze traffic patterns, environmental factors, and driver behavior in order to predict and prevent road accidents. These accidents cause loss of life, injuries, and economic damage. Early identification of accident conditions can mitigate risks, reduce response time, and save lives. We implemented supervised machine learning algorithms. These models were evaluated based on key performance metrics. Experiments were conducted using tools like RapidMiner and Python, and the Random Forest classifier achieved the highest accuracy of 94.8%. This model demonstrates superior predictive capabilities and robust performance in accident detection scenarios. The insights gained from this research contribute to developing intelligent systems for accident prevention, ensuring safer roads and reduced fatalities.
- Open Access
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Predicting Credit Card Scam Using Machine Learning
In the credit card industry, managing customer relationships is essential for reducing customer churn and building long-term loyalty. With the increasing volume of transaction data, machine learning has become a valuable tool for analyzing customer behavior and predicting future actions. This approach helps companies proactively address customer needs, prevent fraud, and tailor services more effectively. The proposed system focuses on improving profitability, reducing fraud through smarter detection, enhancing the user experience, and shifting from a traditional system to a more customer-centered strategy. Several machine learning models including logistic regression, random forest, K-nearest neighbors, and a deep learning model were applied to evaluate performance and determine the most effective approach. The system was tested using a publicly available dataset containing 10,128 records.
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessPredicting Credit Card Scam Using Machine Learning
In the credit card industry, managing customer relationships is essential for reducing customer churn and building long-term loyalty. With the increasing volume of transaction data, machine learning has become a valuable tool for analyzing customer behavior and predicting future actions. This approach helps companies proactively address customer needs, prevent fraud, and tailor services more effectively. The proposed system focuses on improving profitability, reducing fraud through smarter detection, enhancing the user experience, and shifting from a traditional system to a more customer-centered strategy. Several machine learning models including logistic regression, random forest, K-nearest neighbors, and a deep learning model were applied to evaluate performance and determine the most effective approach. The system was tested using a publicly available dataset containing 10,128 records.
- Open Access
- 20 November 2025
- Published by MDPI AG in Preprints.org
Classification of Personality Types Using Predictive Models
Personality prediction is a remarkable field with applications in psychology, marketing, and human-computer interaction. This study used supervised machine learning techniques to predict a person's personality type. The goal of the research was to accurately predict personality types using machine learning algorithms. To achieve this, we applied Gradient Boosting, Decision Tree, Random Forest, and Naive Bayes classifiers. Among these machine learning models, the Random Forest algorithm demonstrated the highest accuracy in predicting personality types, achieving an impressive accuracy rate of 87.64%. This study highlights an effective machine learning approach for personality prediction and offers valuable insights for future research in this domain.
- 20 November 2025
- Published by MDPI AG in Preprints.org
Open AccessClassification of Personality Types Using Predictive Models
Personality prediction is a remarkable field with applications in psychology, marketing, and human-computer interaction. This study used supervised machine learning techniques to predict a person's personality type. The goal of the research was to accurately predict personality types using machine learning algorithms. To achieve this, we applied Gradient Boosting, Decision Tree, Random Forest, and Naive Bayes classifiers. Among these machine learning models, the Random Forest algorithm demonstrated the highest accuracy in predicting personality types, achieving an impressive accuracy rate of 87.64%. This study highlights an effective machine learning approach for personality prediction and offers valuable insights for future research in this domain.
- Open Access
- 20 November 2025
- Published by MDPI AG in Preprints.org
An Advance Machine Learning Technique to Predict Travel Insurance Outcomes
The rapidly growing interest in travel, coupled with the risks associated with it, compels travelers to consider purchasing travel insurance. This research analyzes detailed data on tourists’ choices, interests, and travel history to predict travel insurance behavior. The goal is to uncover patterns and develop a predictive model that assists in identifying the best insurance options for travelers using machine learning algorithms. This study examines feature importance, model performance, and decision-making strategies to effectively target potential customers. Experimental results were obtained using machine learning algorithms such as Decision Tree, Random Forest, KNN, and Naïve Bayes. The effectiveness of ensemble methods and deep learning techniques was also evaluated in achieving high prediction accuracy. Among all the algorithms tested, the Random Forest model achieved the highest accuracy of 83.39%, with a precision of 91.30% and a recall of 96.87%. These findings provide actionable insights for data-driven decision-making in the travel insurance industry, promoting enhanced customer engagement and operational efficiency.
- 20 November 2025
- Published by MDPI AG in Preprints.org
Open AccessAn Advance Machine Learning Technique to Predict Travel Insurance Outcomes
The rapidly growing interest in travel, coupled with the risks associated with it, compels travelers to consider purchasing travel insurance. This research analyzes detailed data on tourists’ choices, interests, and travel history to predict travel insurance behavior. The goal is to uncover patterns and develop a predictive model that assists in identifying the best insurance options for travelers using machine learning algorithms. This study examines feature importance, model performance, and decision-making strategies to effectively target potential customers. Experimental results were obtained using machine learning algorithms such as Decision Tree, Random Forest, KNN, and Naïve Bayes. The effectiveness of ensemble methods and deep learning techniques was also evaluated in achieving high prediction accuracy. Among all the algorithms tested, the Random Forest model achieved the highest accuracy of 83.39%, with a precision of 91.30% and a recall of 96.87%. These findings provide actionable insights for data-driven decision-making in the travel insurance industry, promoting enhanced customer engagement and operational efficiency.
- Open Access
- 20 November 2025
- Published by MDPI AG in Preprints.org
COVID-19 Classification Using Machine Learning
COVID-19 is an infectious disorder caused by the SARS-CoV-2 virus, first identified in December 2019. It can cause a range of symptoms such as pneumonia, fatigue, and even death. The study focuses on the COVID-19 dataset, which contains information about patients, symptoms, test results, recovery, and age details. We apply machine learning algorithms, including Random Forest, Decision Trees, KNN, and Naïve Bayes, to predict outcomes of COVID-19, such as test results, recovery time, and hospitalization probability. The dataset undergoes several processing steps, such as handling missing values, normalization, and feature selection. These algorithms are evaluated to determine accuracy, precision, recall, and the confusion matrix to identify the best model.
- 20 November 2025
- Published by MDPI AG in Preprints.org
Open AccessCOVID-19 Classification Using Machine Learning
COVID-19 is an infectious disorder caused by the SARS-CoV-2 virus, first identified in December 2019. It can cause a range of symptoms such as pneumonia, fatigue, and even death. The study focuses on the COVID-19 dataset, which contains information about patients, symptoms, test results, recovery, and age details. We apply machine learning algorithms, including Random Forest, Decision Trees, KNN, and Naïve Bayes, to predict outcomes of COVID-19, such as test results, recovery time, and hospitalization probability. The dataset undergoes several processing steps, such as handling missing values, normalization, and feature selection. These algorithms are evaluated to determine accuracy, precision, recall, and the confusion matrix to identify the best model.
- Open Access
- 18 November 2025
- Published by MDPI AG in Preprints.org
Machine Learning-Based Approach for Early Prediction of Liver Disease
Artificial intelligence and machine learning are increasingly used to support the diagnosis of liver diseases. They help in diagnosing liver diseases more comfortably, offering alternatives to painful treatments like liver biopsy. This research paper reviews 15 recent studies that discuss how AI and ML improve the diagnosis and accuracy of liver diseases. These studies also highlight improvements in feature selection and data handling. The data on liver diseases is imbalanced, with some conditions having much less data than others, making comparison between studies difficult due to varying methods. Additionally, dynamic datasets that change over time are rarely used, limiting deeper analysis of disease progression. This study highlights how AI and ML can improve liver disease diagnosis by addressing data imbalance, enhancing accuracy, and reducing reliance on invasive procedures.
- 18 November 2025
- Published by MDPI AG in Preprints.org
Open AccessMachine Learning-Based Approach for Early Prediction of Liver Disease
Artificial intelligence and machine learning are increasingly used to support the diagnosis of liver diseases. They help in diagnosing liver diseases more comfortably, offering alternatives to painful treatments like liver biopsy. This research paper reviews 15 recent studies that discuss how AI and ML improve the diagnosis and accuracy of liver diseases. These studies also highlight improvements in feature selection and data handling. The data on liver diseases is imbalanced, with some conditions having much less data than others, making comparison between studies difficult due to varying methods. Additionally, dynamic datasets that change over time are rarely used, limiting deeper analysis of disease progression. This study highlights how AI and ML can improve liver disease diagnosis by addressing data imbalance, enhancing accuracy, and reducing reliance on invasive procedures.