Temporarily lecturer and researcher
Paris 1 Panthéon Sorbonne, 96 hours of teaching: L1 and L2 Level
Bio
Since Spring 2022, I am an associate researcher at Paris 1 Panthéon University and consultant within Experfy-Harvard Innovation Lab. Apart from my mathematical background, I am an expert in industrial risk management and data science with an emphasis on the energy sector.
Academic Positions
-
2020 2021
-
2019 2020
Temporarily lecturer and researcher
Paris 1 Panthéon Sorbonne, 96 hours of teaching: L2 Level
-
2016 2019
Teaching Assistant
Sorbonne University, 192 hours of teaching on various levels from L1 to Master 1.
-
2016 2016
Data Scientist
IFSTTAR, Data mining for profile extraction usage and forecasting in the field of energy.
-
2015 2015
Research engineer
EDF R&D, Functional Data Analysis in an industrial context, analysis of a welding process.
Education & Training
-
Ph.D. 10/2016 - 12/2020
PhD of Applied Mathematics
Sorbonne University
-
Master 09/2015 - 06/2016
Master 2 of Innovation Management
Sorbonne University
-
Master 09/2015 - 06/2016
Master 2 of Mathematics
Sorbonne University
-
Statistician Diploma09/2014 - 09/2016
Statistician Diploma
ISUP: French Grande Ecole
-
Master 109/2013 - 06/2014
Master 1 of Mathematics
Sorbonne University
-
Bachelor 09/2010 - 06/2013
Bachelor of Applied Mathematics
Paris Dauphine University
Awards
-
2018Winner of Data Challenge 2018 organized by EDF and SFDS.2016Valedictorian at Sorbonne University among 55 students, the only student to publish two articles before starting the PhD.2014Youngest contributor in the R Package Sensitivity in 2014.2015Youngest Poster Presenter during the Annual conference of the GDR MASCOT NUM 2015.
Research Summary
Touati's current research revolves around probability theory, statistics, machine learning and renewable energy engineering.
Interests
- Applied Mathematics
- Concentration inequalities
- Martingales
- Elephant and Minimal Random Walks
- Online learning
- Engineering
- Renewable Energy
- Energy Analytics
- Sensitivity Analysis
Publications
Filter by type:
Sort by year:
New insights on concentration inequalities for self-normalized martingales
PublicationDOI:10.1214/19-ECP269
Abstract
We propose new concentration inequalities for self-normalized martingales. The main idea is to introduce a suitable weighted sum of the predictable quadratic variation and the total quadratic variation of the martingale. It offers much more flexibility and allows us to improve previous concentration inequalities. Statistical applications on autoregressive process, internal diffusion-limited aggregation process, and online statistical learning are also provided.
Hourly Solar Irradiance Forecasting Based on Machine Learning Models
PublicationDOI:10.1109/ICMLA.2016.0078
Abstract
In recent years, many research studies are conducted into the use of smart meters data for developping decision-making tools including both analytical, forecasting and display purposes. Forecasting energy generation or forecasting energy consumption demand are indeed central problems for urban stakeholders (electricity companies and urban planners). These issues are helpful to allow them ensuring an efficient planning and optimization of energy resources. This paper investigates the problem for forecasting the hourly solar irradiance within a Machine Learning (ML) framework using Similarity method (SIM), Support Vector Machine (SVM) and Neural Network (NN). These approaches rely on a methodology which takes into account the previous hours of the predicting day and also the days having the same number of sunshine hours in the history. The study is conducted on a real data set collected on the Paris suburb of Alfortville. A comparison with two time series approaches namely Naive method and Autoregressive Moving Average Model (ARMA) is performed. This study is the first step towards the development of the hourly solar irradiance forecasting hybrid models.
Abstract
In this communication, the extension of the Martinez method to non Gaussian distribution is studied. Indeed, non Gaussianity can distort the Fisher’s confidence interval, and the outcome can be quite misleading. The two following points will be discussed, the R implementation of the method is available in the Package Sensitivity).
Teaching
List of courses
-
2M261 77 Hours
Integer series, integral depending on a parameter, applications to differential equations
Sorbonne Université. Bachelor of Mathematics, second Year.
-
3M100 21 Hours
Initiation to Python
Sorbonne Université. Bachelor of Mathematics, Third Year.
-
1M001 38 Hours
Algebra and Calculus for Science
Sorbonne Université. Bachelor of Science, First Year.
-
4M015 24 Hours
Statistics.
Sorbonne Université. Master of Mathematics, First Year.
-
1M... 21 Hours
Mathematical Statistics.
Sorbonne Université. Master of Actuary, First Year.
-
3M101 10 Hours
Applied Mathematics Project.
Sorbonne Université. Bachelor of Mathematics, Third Year.
-
2M... 96 Hours
Numerical Methods.
Sorbonne Université. Bachelor of Applied Mathematics, Second Year.
-
2M... 96 Hours
Calculus 1 and Real Analysis
Paris 1 Sorbonne. Bachelor of Applied Mathematics, First and Second Year.
-
Math243 42 Hours
Discrete Mathematics.
MedTech. Freshman and Sophomore Level.
-
Math244 Hours
Probability and statistics.
MedTech. Freshman and Sophomore Level.
Certificates
I have obtained following certificates.
Certificates
- Reinforcement Learning Coursera Certificate: University of Alberta (In preparation)
- Advanced Data Science with IBM Coursera Certificate ( In preparation)
Contact
The best way to reach me is by email.
At My Office
C20-10, at the SAMM Department
Université Paris I, Panthéon Sorbonne
Centre PMF
90 Rue de Tolbiac
75013, Paris, France.