M. MOHAMED Djerioui

Alzheimer’s disease (AD) is a chronic, progressive neurodegenerative disorder that typically affects elderly individuals. Detecting Alzheimer’s using plasma proteins is a critical step toward improving treatment results for this disease. This study aims to use computational algorithms to explore the relationship between plasma proteins and AD progression by identifying a panel of plasma proteins that can serve as biomarkers for tracking and diagnosing AD. We applied two feature selection methods, Sequential Backward Feature Selection (SBFS) and Analysis of Variance (ANOVA) to extract significant proteins from a dataset of 146 proteins. The data was collected from the plasma of 566 individuals, comprising both Alzheimer’s patients and healthy controls. The SBFS technique generated all possible combinations of protein groups from the 146 proteins, which were then trained and tested using five machine learning models: Decision Tree, Random Forest, Extremely Randomized Trees, Extreme Gradient Boosting, and Adaptive Boosting. Subsequently, ANOVA was applied to refine and reduce the selected panel size. Finally, we used XGBoost and AdaBoost models to validate the final panel. The findings introduce a plasma protein panel consisting of A2Macro, BNP, BTC, PPP, and PYY proteins for diagnosing AD. This panel achieved a sensitivity of 88.88%, a specificity of 66.66%, and an AUC of 0.85. These results demonstrate that plasma protein biomarkers can facilitate timely interventions, potentially slowing disease progression and improving patient outcomes. This non-invasive and affordable diagnostic method has the potential to make Alzheimer’s screening accessible to a broader population.

Alzheimer’s disease (AD) is a gradient degeneration of essential cognitive activities such as memory, thinking, and cognition. AD mainly affects elderly individuals and is recognized as the most common cause of dementia. This study investigates the predictive performance of nine supervised machine learning algorithms—Logistic Regression, Decision Tree, Random Forest, K-Nearest Neighbors, Support Vector Machine, Gaussian Naïve Bayes, Multi-Layer Perceptron, eXtreme Gradient Boost, and Gradient Boosting—using neuropsychological assessment data. We applied two classification techniques—binary and multiclass—to classify 1761 subjects into three categories: cognitively normal (CN), mild cognitive impairment (MCI), and Alzheimer's disease (AD). Binary classification tasks focused on CNvsAD and CNvsMCI subsets, while multiclass classification used the full dataset (TriClass). Hyperparameter tuning was performed to optimize model performance. The results indicate that ensemble learning models, particularly Gradient Boosting (GB) and Random Forest (RF), exhibited superior accuracy compared to other algorithms. Most models for the CNvsAD subset achieved the highest accuracy (97.74%), while GB achieved the best performance (94.98%) for the CNvsMCI subset. For multiclass classification, RF achieved the highest accuracy at 84.70%. These findings highlight the robustness and efficiency of ensemble learning algorithms, especially in handling complex, non-linear data structures. This study underscores the potential of RF and GB as reliable tools for early detection and classification of Alzheimer’s disease using neuropsychological data.

Diabetic retinopathy (DR) presents a significant threat to global visual health. This
study introduces a method leveraging deep learning and machine learning to
enhance DR detection accuracy. By combining pre-trained CNNs (MobileNetV2,
ResNet50, and Xception) with high performance classifiers like SVM and KNN, our
method significantly improves diagnostic performance. The combination of
ResNet50 and SVM achieved a 95.90% accuracy in detecting retinal abnormalities
for the APTOS 2019 Blindness Detection dataset, demonstrating its superiority over
current diagnostic techniques.

In this paper, we propose a new strategy to exploit the advantages of Deep Neural Network-based architectures for brain tumor classification using MRI images for a better diagnosis. This was achieved by analyzing and evaluating pre-trained models on three different datasets. To better design the optimal architecture for solving the classification of brain tumor using MRIs, we have conducted extensive experiment-based analysis on how different layers of Convolutional Neural Network (CNN) process the inputs. Four distinct architectures are then built, each with its specific hyperparameters and layers. The images are fed into the convolutional layers for feature extraction followed by a softmax function before applying the classification process. An extensive experimental study carried out clearly demonstrates that our novel CNN-based classification approach achieves state-of-the-art accuracy, precision, recall and an F1-score of 99.76% 99.64% 99.62% and 99.64%, respectively. Also, a higher performance in terms of Micro-Avg Matthew correlation coefficient (MCC) of 0.929 is achieved. This exceptional performance is achieved thanks to the new proposed model's architecture. Indeed, unlike conventional methods, that often rely on complex transfer learning models or hybrid architectures, our approach utilizes a custom and non-hybrid scheme. Consequently, this streamlined architecture offers a significant advantage of being remarkably lightweight, enabling efficient operation on resource-constrained computing systems.

Alzheimer's disease (AD) is a progressive neurodegenerative disease with vast worldwide consequences. Timely and accurate diagnosis is crucial but present methods are costly and resource-intensive. This research work investigates the use of CSF and plasma biomarkers, specifically amyloid beta ( ) and tau, in conjunction with machine learning (ML) and deep learning (DL) algorithms to enhance AD prediction. The proposed method seeks to close the gap in accurate early detection and classification. The findings show that improved computational techniques hold the potential to enhance AD diagnosis by leveraging CSF and plasma biomarkers.

Diabetes mellitus is a leading cause of diabetic retinopathy (DR), which results in retinal lesions and vision impairment. Untreated DR can lead to blindness, highlighting the need for early diagnosis and treatment. Unfortunately, DR has no cure, and treatments only help to preserve vision. Traditional manual diagnosis of DR retina fundus images by ophthalmologists is time-consuming, costly, and prone to errors. Computer-aided diagnosis methods, such as deep learning, have emerged as popular methods for improving diagnosis and reducing errors. Over the past decade, Convolutional Neural Networks (CNNs) have been shown to perform very well in medical image analysis due to their high ability to extract local features from images. Convolutional neural networks (CNNs) have shown great success in the processing of medical images, including DR color fundus images. In this paper, we proposed a multi-level fine-tuned deep learning based approach for the classification of diabetic retinopathy using three different pre-trained models including: DenseNet121, MobileNetV2, and Xception. The results are provided as classification accuracy, loss metrics, and the performance is compared with state-of-the-art works. The results indicates that the proposed Xception network surpassed its peers’ models as well as state-of-the-art methods by achieving the highest accuracy of 97.95% in binary classification of DR images.

Monitoring of water quality is one of the world’s main intentions for countries. Classification techniques based on support vector machines (SVMs) and artificial neural network (ANN) has been widely used in several applications of water research. Water quality assessment with high accuracy and efficiency with innovational approaches permitted us to acquire additional knowledge and information to obtain an intelligent monitoring system. In this paper, we present the use of principal component analysis (PCA) combined with SVM and ANN with decision templates combination data fusion method. PCA was used for features selection from original database. The multi-layer perceptron network (MLP) and the one-against-all strategy for SVM method have been widely used. Decision templates are applied to increase the accuracy of the water quality classification. The specific classification approach was employed to assess the water quality of the Tilesdit dam in Algeria as a study area, defined with a dataset of eight physicochemical parameters collected in the period 2009–2018, such as temperature, pH, electrical conductivity, and turbidity. The selection of the excellent parameters of the used models can be improving the performance of classification process. In order to assess their results, an experiment step using collected dataset corresponding to the accuracy and running time of training and test phases, and robustness to noise, is carried out. Various scenarios are examined in comparative study to obtain the most results of decision step with and without feature selection of the input data. From the results, we found that the integration of SVM and ANN with PCA yields accuracy up than 98%. The combination by decision templates of two classifiers SVM and ANN with PCA yields an accuracy of 99.24% using k-fold cross-validation. The combination data fusion enhanced expressively the results of the proposed monitoring framework that had proven a considerable ability in surface water quality assessment.

Breast cancer has become a major leading cause of death and incapacity worldwide. Recently, breast cancer is being responsible for a huge number of deaths of the female gender. In this study, we have implemented the Twin-Support Vector Machine (TW-SVM) to illustrate the power of machine learning techniques. TW-SVM is a recently developed algorithm and yet it is very powerful. For performance measurement, a competitive comparison between the proposed TW-SVM and SVM classifiers has been done based on the WDBC dataset. The results showed that TW-SVM can provide promising performance rates. It outperformed the SVM algorithm as well as other existing works by achieving the highest accuracy of 99.11% for predicting the considered disease.

This paper studies the application of soft-sensing modeling approach for monitoring surface water quality with artificial intelligence techniques such as SOM (Self-Organizing Maps of Kohonen) and SVM (Support Vector Machines). In the water treatment process, many monitoring parameters are expensive or difficult to measure in real-time, limiting the possibilities for highly efficient control of the water production process. In this work, an intelligent soft-sensor was developed to predict optimal coagulant dosage. It confirmed that the coagulation-flocculation unit is essential in producing drinking water. The soft sensor proposed in this paper contains SOM in feature selection and the SVR method for predicting the optimal coagulant dose values. The surface water quality from Mostaganem's Cheliff Dam was advanced assessed (Algeria). The water quality assessment successfully demonstrated the proposed approach's performance and efficacy, and it can achieve complete expertise in the study area.

A brain tumor is a fatal disease that affects children and adults The disease might be detected using a physical exam or a neurological exam, but for the classification, it is done with biopsy. That last one is concerned with brain surgery, which is very hard and complicated in itself. Early detection and classification could help to choose the perfect plan for treatment. With the great development and change in technology, DL techniques could help in diagnosis and classification without any huge risks. Using the available data of Magnetic Resonance Imaging (MRI), that is studied by the radiologist, In our study, we took two approaches, the first including a transfer learning model and the second including a Convolutional Neural Network (CNN) model, to both classify different types of brain tumors. With the CNN approach, we managed to achieve an accuracy of 90 %. The experimental results show that our proposed CNN gives the best accuracy as compared to the transfer learning model.

Transfer learning approche for alzhimer's dicease diagnosis using mri image

One of the main reasons for disability and premature mortality in the world is diabetes disease, which can cause different sorts of damage to organs such as kidneys, eyes, and heart arteries. The deaths by diabetes are increasing each year, so the need to develop a system that can effectively diagnose diabetes patients becomes inevitable. In this work, an efficient medical decision system for diabetes prediction based on Deep Neural Network (DNN) is presented. Such algorithms are state-of-the-art in computer vision, language processing, and image analysis, and when applied in healthcare for prediction and diagnosis purposes, these algorithms can produce highly accurate results. Moreover, they can be combined with medical knowledge to improve decision-making effectiveness, adaptability, and transparency. A performance comparison between the DNN algorithm and some well-known machine learning techniques as well as the state-of-the-art methods is presented. The obtained results showed that our proposed method based on the DNN technique provides promising performances with an accuracy of 99.75% and an F1-score of 99.66%. This improvement can reduce time, efforts, and labor in healthcare services as well as increasing the final decision accuracy.

One of the most leading death causes in the world is brain tumor. Solving brain tumor segmentation and classification by relying mainly on classical medical image processing is a complex and challenging task. In fact, medical evidence shows that manual classification with human-assisted support can lead to improper prediction and diagnosis. This is mainly due to the variety and the similarity of tumors and normal tissues. Recently, deep learning techniques showed promising results towards improving accuracy of detection and classification of brain tumor from magnetic resonance imaging (MRI). In this paper, we propose a deep learning model for the classification of brain tumors from MRI images using convolutional neural network (CNN) based on transfer learning. The implemented system explores a number of CNN architectures, namely ResNet, Xception and MobilNet-V2. This latter achieved the best results …

Heart disease is the leading cause of death in the world according to the World Health Organization
(WHO). Researchers are more interested in using machine learning techniques to help medical staff diagnose
or detect heart disease early. In this paper, we propose an efficient medical decision support system based on
twin support vector machines (Twin-SVM) for heart disease diagnosing with binary target (i.e. presence or
absence of disease). Unlike conventional support vector machines (SVM) that finds only one optimal hyper-
plane for separating the data points of first class from those of second class, which causes inaccurate decision,
Twin-SVM finds two non-parallel hyper-planes so that each one is closer to the first class and is as far from
the second class as possible. Our experiments are conducted on real heart disease dataset and many evaluation
metrics have been considered to evaluate the performance of the proposed method. Furthermore, a
comparison between the proposed method and several well-known classifiers as well as the state-of-the-art
methods has been performed. The obtained results proved that our proposed method based on Twin-SVM
technique gives promising performances better than the state-of-the-art. This improvement can seriously
reduce time, materials, and labor in healthcare services while increasing the final decision accuracy.

One of the key causes of premature disability and mortality in the world today is heart disease, which makes its prediction a vital problem in the field of healthcare systems. This work provides a contribution to the study and creation an intelligent system based on LSTM technique for heart disease prediction. A comparative study is presented between Multi Layer Perceptron (MLP) and Long Short Term Memory (LSTM) techniques in terms of accuracy and other predictive parameters for heart disease. The main aim is to develop an intelligent system based on LSTM technique for predicting heart disease in order to make an adapted decision to prevent and monitor heart disease and stroke. As it has better characteristics than those of the MLP technique, LSTM is shown to be the most effective technique for solving the aforementioned problems.

Actuellement, dans le domaine de traitement des eaux propres, l’utilisation de procédés
automatiques devient impérative pour atteindre deux objectifs principaux : la maîtrise de la qualité de l’eau, et
la diminution des contraintes de coût de fonctionnement. Les variables de qualité jouent un rôle important dans
le contrôle et la surveillance de l'eau. Bien que certains paramètres puissent être mesurés en continu à l'aide de
capteurs physiques à faible coût, il existe d'autres paramètres nécessitant des analyses de laboratoire spécifiques
et coûteuses en raison de l'absence de capteurs dédiés. En outre, il existe un grand nombre de capteurs
hétérogènes qui peuvent prendre beaucoup de temps lors des étapes de mesure et de traitement. Néanmoins, le
système à base des capteurs logiciels peut fournir un moyen efficace et économique pour résoudre ces
problèmes.
Notre travail peut être considéré comme une contribution aux solutions proposées, pour résoudre des
problèmes d’intérêt stratégique à préoccupation nationale et internationale, utilisant des outils modernes à base
de techniques avancées. On présente dans ce cadre, une contribution à l'étude et au développement de capteurs
logiciels pouvant être intégrés au niveau de capteurs intelligents et utilisés dans la surveillance et le contrôle de
la qualité des eaux propres. Ces capteurs algorithmiques à base de techniques de l’intelligence artificielle et
représentant un moyen très attrayant pour faire face au manque de capteurs spécifiques, sont devenus alors très
convoités. Dans un but de choix de la technique la mieux adaptée à la conception de ces capteurs logiciels, une
étude comparative d’évaluation de plusieurs méthodes, est à cet effet présentée. L'objectif est de mettre en
œuvre une architecture de système de surveillance fondée sur l’emploi de ce type de capteurs, pouvant être
intégrés au sein de plateformes plus élaborées, construites autour de capteurs intelligents.

Water quality monitoring are fundamental tools in the management of water resources and they provide essential information characterizing status of water resources, determining trends and changes over time, and identifying emerging water quality issues. This task consists of collecting quantitative information of different water parameters through a statistical sampling. For each parameters measurement, a sensor or a specific treatment is made. This makes the quality monitoring very expensive in money, time and labor. Therefore, it is important that water quality issues need to be understood in the framework of hydrological processes based on the water quality and hydrological monitoring.
To remedy this problem, we propose in this work an efficient system for water quality classification using Minimum Redundancy Maximum Relevance (mRMR) and Extreme Learning Machine (ELM). The mRMR is an algorithm frequently used in a method to accurately identify characteristics to reduce the number of input water parameters introduced in the ELM classifier and is usually described in its pairing with relevant feature selection. These subsets often contain material which is relevant but redundant and mRMR attempts to address this problem by removing those redundant subsets. mRMR has a variety of applications in many areas such as pattern recognition. As a special case, the "correlation" can be replaced by the statistical dependency between variables. Mutual information can be used to quantify the dependency. In this case, it is shown that mRMR is an approximation to maximizing the dependency between the joint distribution of the selected features and the classification variable. The ELM which is a technique for pattern classification has been widely used in many application areas such as water quality monitoring. A multi-class problem using ELM is a typical example for solving the mentioned problem.
In this work, Experimental results conducted on real dataset collected from Tilesdit dam of Bouira state (Algeria) were selected for this study. The proposed feature selection method can efficiently reduce the number of water parameters needed to classify its quality, which consequently causes a minimization in number of required sensor/treatment. Therefore, the water quality classification is perfectly insured with the trade-off between the low-cost and a high accuracy. Its performance is more competitive when compared with artificial neural networks. Furthermore, the results demonstrated that the proposed procedure has a great potential in water quality monitoring.

In this investigation a novel type-2 fuzzy model for Time series is presented. It is based on interval type-2 fuzzy systems. The proposed method deals with the curve fitting and computational time problems of type-2 fuzzy systems. This approach will significantly reduce the number of type-2 fuzzy rules and simultaneously preserves the fitting quality. The proposed model comprises a parallel interconnection of two type-2 sub-fuzzy models. The first one is the primary model, which represents an ordinary model with a low resolution for the time series under consideration. To overcome resolution quality problem and obtain a model with higher resolution, we introduce the second model called the error model. This model represents the error modelling between the primary model and the real time series model. The error model characterizes the uncertainty in the primary model which can be minimized by a simple subtraction of the error model output from the primary model output. The result is a parallel interconnection between the two sub models. Thus, a unique and entire final model possessing higher resolution is realized. The model's representation and identification are implemented by using type-2 fuzzy auto regressive moving average (T2FARMA) models. Identification is achieved by innovative metaheuristic optimization algorithm such as biogeography-based optimization (BBO). The effectiveness of the method is evaluated by testing the proposed model with the reference time series models. In addition, a detailed comparative study with several reference methods will be presented. The results of the experiments that have been conducted confirm that the proposed method can considerably improve convergence, resolution and computation time.

In this paper, we propose a heart disease prediction system based on Neighborhood Component Analysis (NCA) and Support Vector Machine (SVM). In fact, NCA is used for selecting the most relevant parameters to make a good decision. This can seriously reduce the time, materials, and labor to get the final decision while increasing the prediction performance. Besides, the binary SVM is used for predicting the selected parameters in order to identify the presence/absence of heart disease. The conducted experiments on real heart disease dataset show that the proposed system achieved 85.43% of prediction accuracy. This performance is 1.99% higher than the accuracy obtained with the whole parameters. Also, the proposed system outperforms the state-of-the-art heart disease prediction.

Multimodal biometrics is an evolving technology in the fields of security. Biometrics system reduces the effort of remember a memorable password. Multimodal biometrics system uses two or more traits for efficient recognition. This paper presents a hand biometric system by fusing information of palmprint and finger knuckle. To this end, BSIF ( Binarized Statistical Image Features) filter and LBP (Local binary patterns) coefficients are employed to obtain the Finger-knuckle-print and palm-print traits, and subsequently selection of the features vector is conducted with PCA (Principal Component Analysis) transforms in higher coefficients. To match the finger knuckle or palm-print feature vector, the (ELM) Extreme learning machine is applied. According to the experiment outcomes, the proposed system not only has a significantly high recognition rate but it also affords greater security compared to the single biometric system.

Nowadays, one of the main reasons for disability and mortality premature in the world is the
heart disease, which make its prediction is a critical challenge in the area of healthcare systems.
In this paper, we propose a heart disease prediction system based on Neighborhood Component
Analysis (NCA) and Support Vector Machine (SVM). In fact, NCA is used for selecting the
most relevant parameters to make a good decision. This can seriously reduce the time,
materials, and labor to get the final decision while increasing the prediction performance.
Besides, the binary SVM is used for predicting the selected parameters in order to identify the
presence/absence of heart disease. The conducted experiments on real heart disease dataset
show that the proposed system achieved 85.43% of prediction accuracy. This performance is
1.99% higher than the accuracy obtained with the whole parameters. Also, the proposed system
outperforms the state-of-the-art heart disease prediction.

A major problem in water treatment plants is the continuous difficulty faced in online measurement by means of dedicated measuring hardware and laboratory analysis of certain variables related to the composition of water. Actually, for several reasons, such as the high cost of some sensors, their number, the dedicated time to check out the sensors, cleaning operation, calibration routines and sensor replacement, make their proper operation hard to ensure high-quality composition of water. Furthermore, in water quality monitoring, there is a huge number of heterogeneous sensors which may be time-consuming in the measurement and processing stages. Nevertheless, soft sensor approach can provide an effective and economic way to solve this problem for any cases of sensor failure. This work presents a contribution to the study and development of a soft sensor used in water quality monitoring using chlorine. A comparative study between support vector machine (SVM) and extreme learning machine (ELM) techniques in terms of learning time and other parameters for regression and classification is presented. The main objective is to set up a system architecture based on a soft sensor for water quality in order to make an adapted decision to the control and monitoring of water quality issues. ELM is shown to be the most suitable technique to address the previously mentioned problems as it has better characteristics than those of the SVM technique. An example of application is provided to focus on the interest of using a chlorine soft sensor as it is accurate, efficient and less cost-effective tool.