The aim of this study was to determine, using Machine Learning (ML) algorithms, whether a pregnant woman will give birth by caesarean or not. The study is based on quantitative analysis using secondary data from the obstetric department of the Wanamahika Hospital in the city of Butembo, the Democratic Republic of Congo, over a period of one year and seven months in 2019 and 2020. The used dataset comprises 1501 records. Six ML models, namely: the Decision Tree, the Support Vector Machines, the Artificial Neural Networks, the k-Nearest Neighbors, the Random Forest as well as the Logistic Regression (LR), were built to predict the mode of women deliveries. The evaluation metrics used to evaluate those models were accuracy, f1-score, precision and recall. The authors found that cesarean deliveries represented 33.8% of their study sample while vaginal deliveries represented 66.2%. Of those six ML models created, LR was validated as it performed better with an accuracy reaching 98.85%, a recall, a precision, and a f1-score of 0.99, respectively. At the end of the investigation, the researchers retained LR in order to deploy a Web application that detect the mode of women deliveries in the hospital using Flask. Nineteen features revealed to be predictors of delivering in caesarean mode in the City of Butembo that are Referred by the health center, Age of the mother, Origin of the mother, Age of the pregnancy, HIV test result, Antepartum hemorrhage, Uterine rupture, Eclampsia and pre-eclampsia, Active management of the third period of labor, Indications for cesarean section, Number of previous cesarean sections, Episiotomy, Theobald, Cytotec, Sex of live birth, Obstetric formula, Weight of newborn in grams, Number of days in hospital, Number of days before delivery.

Heart diseases are the leading cause of death worldwide. Nowadays, heart diseases are killing more people than ever before. Thus, the authors of this study designed this project to analyze data on heart diseases prediction. The project uses raw data in the form of a.csv file as data set. The authors collected the used dataset from the cardiology department of the Graben University Clinic (Butembo/DR Congo) that included 389 records and 25 variables including age, employment, pulse rate, blood pressure and clinical symptoms. The aim was to compare Machine Learning (ML) ensemble methods such as Boosting type (AdaBoosting, GradientBoosting and XGBoosting) with single ML models (KNN, Stochastic gradient descent (SGD), Decision Tree) to see which of the models predict better heart diseases in unstable and insecure areas. Thus, the results showed that the XGBoost model performed better with accuracy, precision and recall of 85% respectively. In this research the authors concluded that Boosting as ensemble method classifies accurately heart diseases data in an insecure area such as Butembo, in the province of North-Kivu, DR Congo.

Training a multi-layer neural network is sometimes difficult, especially for novices in Artificial Intelligence. Many people believe that this training must be relayed to computers in order to be able to perform these ultra-powerful calculations. As a result, we can't figure out what is going on behind these calculations, thinking that there is too much mathematics, making it difficult for humans to understand what is at stake. Far from this mythical caricature stuck to neural networks. The training of a neural network consists in finding synaptic weights such that the output layer allows to classify with precision the observed values of a training set with the aim of allowing the created model to present generalisation capacities on examples that it will never have encountered during the training phase.

E-commerce is distinguished from traditional commerce through the dematerialization of its activities due to the use of Information and Communication technologies over the Internet. This research examines issues related to the problems of dematerialization of e-commerce activities, as well as its omnipresence (ubiquity) manifested by the fact that the e-commerce website is accessible in almost all countries and depersonalization, which creates risks due, on the one hand, to the lack of the physical presence of the contractors and, on the other hand, to the use of the electronic medium to conclude the contract. The objective is to provide information on the applicability of virtual commerce practices in order to secure the professional environment of the e-commerce through the protection of general rules such as information provided by companies, practices unfair commercial terms, unfair contract clauses, online payment security, data protection and confidentiality, dispute resolution and remedies, and international electronic transactions.