Alexis Huet Email & Phone Number

Paris Area, France

Nanjing City, China

Nanjing Howso Technology is an IT company providing optimization services for mobile operators (China Mobile, China Telecom) and network operators (Ericsson, ZTE, Huawei).My role was to organize and analyze large amounts of data from telecom sources. I was responsible for data mining, modeling and implementation for different machine learning projects. The purpose of these projects was mainly to solve classification, clustering, forecasting and optimization tasks. They lead to scientific articles, technical reports and patent applications.Technical environment: R, Python.Algorithms: - visualization and dimensionality reduction techniques (t-SNE, PCA…),- deep learning methods (MLP, CNN, RNN),- regression (GLM, GAM…), - classification (random forest, decision trees, SVM…), - clustering (GMM, ensemble clustering…),- time series (ETS, ARIMA…).Example of projects in production:Project: LTE service performance for high-speed rail cells based on signaling data- Modeling and implementation in R (processing using dplyr, random forests with caret),- Presentation at ICSINC 2017 (Chongqing, China),- Improving classification results using Python/Keras,- Visualization with t-SNE.Project: Latent clustering models for outlier identification in telecom data- Implementation in R (preprocessing, clustering model, ggplot outputs),- Article in Hindawi MIS.Project: Coverage and capacity optimization for 4G LTE networks using Differential Evolution- Reorganizing existing code (improved modularity, wrote documentation),- Speed optimization (vectorization, parallelization, use of existing packages).

Lyon Area, France

I did my PhD thesis in applied mathematics under the supervision of Anne-Laure Fougères and Jean Bérard. The aim of the thesis was to approximate likelihoods in biological evolutionary models where dependence phenomena of long range is present. To this end, Markov structures have been identified, making the use of Sequential Monte Carlo methods possible.General workflow:- Updating hidden Markov structures,- Using sequential Monte Carlo methods,- Implementation in C++,- Applications on both simulated and biological sequences.PhD thesis available at: https://www.theses.fr/en/2014LYO10126

Lyon Area, France

Teaching assistant for bachelor’s degree students. Overall volume: 288 hours.Courses included linear regression, analysis of variance, probability, integration and differential calculus; with practical exercises in R.

Doctor Of Philosophy - Phd, Applied Mathematics

My PhD thesis is available at this address: https://www.theses.fr/en/2014LYO10126

Master Of Science - Ms, Mathematical Statistics And Probability

Agrégation Of Mathematics

In France, the agrégation is a competitive examination for civil service in the French public education system. Subjects for 2009 session.

Research Internship, Mathematics

First Year Of Master, Mathematics

Bachelor'S Degree, Mathematics

Classe Préparatoires Mpsi/Mp*

Baccalauréat Scientifique