Sidy Mamadou Diallo Email and Phone Number

Modeling the environmental impact of telecom networks usingartificial intelligence

Purpose: To select the best antibodies.Preprocessing: - Normalization - Missing Data - Dimension reduction (PCA, t-SNE, LLE) - Neighborhood detection (DBSCAN, Isolation Forest)Clustering: -PAM -Hierarchical clustering - Gaussian mixture modelRshiny (shinydashboard, plotly, tidyverse, git, gitlab, jenkins CSS, HTML, JS, etc) to automate the process.

Regression and sensitivity analysis using machine learning methods.Physical phenomenon: Interaction between fuel (corium) and coolant (water) during steam explosion.Experimental designMachine learning method: -Random forest, Stacking, boosting -Neural networks -Gaussian processesSensitivity analysis: Sobol index and Random ForestValidation:Evaluate whether the implemented learning methods arecompatible with the physical constraints of the underlying modelsthat provided the dataProgramming languages: Python, Uranie and R