The fact that the signature is widely used as a means of personal verification emphasizes the need for an automatic verification system because of the unfortunate side-effect of being easily abused by those who would feign the identification or intent of an individual. A great deal of work has been done in the area of off-line signature verification over the past few decades. Verification can be performed either Offline or Online based on the application. Online systems use dynamic information of a signature captured at the time the signature is made. Offline systems work on the scanned image of a signature. In this paper, we present a method for Offline Verification of signatures using a set of simple shape based geometric features. The features that are used are Area, Center of gravity, Eccentricity, Kurtosis and Skewness. Before extracting the features, preprocessing of a scanned image is necessary to isolate the signature part and to remove any spurious noise present. The system is initially trained using a database of signatures obtained from those individuals whose signatures have to be authenticated by the system. The details of preprocessing as well as the features depicted above are described throughout the discussion. Then artificial neural network (ANN) was used to verify and classify the signatures: exact or forged, and a classification ratio of about 93% was obtained under a threshold of 90%.The implementation details and simulation results are discussed in the thesis.

Forest fires are among the most dangerous natural threats that bring calamities to a community and can turn it totally upside down.In this paper, to enable a prevention mechanism, we rely on analytics to build a novel fire danger index model that predicts the risk of a developing fire in north Lebanon. We use correlation methods such as statistical regression, Pearson, Spearman and Kendall's Tau correlation to identify the most affecting parameters on fire ignition during the last six years in north Lebanon.The correlations of these attributes with fire occurrence are studied in order to develop the fire danger index. The strongly correlated attributes are then derived. We rely on linear regression to model the fire index as function of a reduced set of weather parameters that are easy to measure.This is critical as it facilitates the application of such prevention models in developing countries like Lebanon. The outcomes resulting from validation tests of the proposed index show high performance in the Lebanese regions. An assessment versus common widespread weather models is then made and has showed the significance the selected parameters.It is strongly believed that this index will help improve the ability of fire prevention measures in the Mediterranean basin area.

<b><i>Objective:</i></b> Pre-eclampsia (PE) is a serious disease of pregnancy and one of the major causes of morbidity and mortality for both the mother and baby. This systematic review aims to detect the role of high-sensitivity C-reactive protein (CRP) in the detection of PE. <b><i>Methods:</i></b> Thirty-four articles published between 2001 and 2019 were included in this review. The articles were extracted from OVID Medline and Embase. The study designs of these articles are randomized controlled, cohort, case-control, and cross-sectional studies evaluating CRP as a marker to predict or early diagnose PE. The quality assessment of these articles is made by the modified Quality Assessment of Diagnostic Accuracy Studies 2 tool. Meta-analysis was not done because of clinical and statistical heterogeneity. <b><i>Results:</i></b> A positive association between CRP levels and the development of PE was confirmed in 18 studies. This positive effect was addressed in patients with normal BMI (&#x3c;25 kg/m²) in 3 studies and in overweight patients in 2 studies. One study addressed this positive association in patients with a BMI ranging between 28 and 31 kg/m². Three studies determined a cutoff level of CRP above which a significant risk of PE development should be suspected. These levels ranged between 7 and 15 mg/L. <b><i>Conclusion:</i></b> CRP is a promising cost-effective biomarker that may be used in the prediction of PE. A CRP level higher than 15 mg/L may suggest initiation of low-dose aspirin in low-risk pregnancies.

Lebanon is known as a tourist destination for its scenic green mountains but the fires have been threatening this green forestry all over the world. The consequences of forest fires are disastrous on the natural environment and ecological systems, not to mention the population, by worsening poverty and lowering the quality of life. Two data mining techniques are used for the purpose of prediction and decision-making: Decision trees and back propagation forward neural networks. Four meteorological attributes are utilized: temperature, relative humidity, wind speed and daily precipitation. The obtained tree drawn from applying the first algorithm could classify these attributes from the most significant to the least significant and better foretell fire incidences. Adopting neural networks with different training algorithms shows that networks with 2 inputs only (temperature and relative humidity) retrieve better results than 4-inputs networks with less mean squared error. Feed forward and Cascade forward networks are under scope, with the use of different training algorithms.

Background: Road accidents have become a major social and health problem as they have dramatically increased worldwide. The statistical study of the given data has revealed that there are no correlations among the various attributes, so it is worth using the most advanced techniques in data analysis and data mining techniques. Our findings serve to develop preventive policies and measures that contribute to promote the concept of road safety. Methods: In this paper, we have applied an association rule mining technique, Apriori Algorithm, to interpret the relationships among causal factors of road accident. Outranking problem has taken place as the Apriori Algorithm has extracted a huge number of rules in which there is a difficulty for decision makers to make the preference decision according to the most interesting rules. The integration of MCDA methods, ELECTRE TRI and PROMETHEE, has then been promoted specifically on set of rules. Results: The application of MCDA methods, ELECTRE TRI and PROMETHEE, has resulted in the same number of the most relevant rules according to the third category C3 recording the highest weight among criteria. ELECTRE TRI has listed the 10 most interesting rules, while the PROMETHEE method has ranked the same 10 rules in terms of the preference between the alternatives and the criteria. We have found that fatal road accidents in Lebanon occur mostly in the following cases: -The most common and critical cause of fatal RTA is speeding, and mainly in the case of rollover accidents. Speed has its direct impact on both the risk of a car accident and its consequences. -Fatal RTA frequently takes place when the weather is rainy and the road is wet. -Fatal RTA considerably occurs when pedestrians are one party of the accidents. -Fatal RTA often occur when we have no official holidays and the road is wet and particularly in single car accidents Conclusion: Upon viewing the results, we can conclude that a driver’s behavior influences the occurrence of all RTA. Moreover, the results have shown the integration of the utilized MCDA methods has led to meaningful information that serve to develop and adopt preventive policies and measures that will improve road safety