M2 Choice – Statistics and Econometrics

M2EcoStat_currentstudent2Current Student – Joseph Agossa

Which aspects of your chosen program were the most challenging ?

Among the courses I have chosen this year, I can say that the most challenging for me were Mathematics of deep learning algorithms. Deep learning knowledge can be described in terms of four distinct aspects:

  • Knowledge of multiple models and multiple viewpoints of the domain.
  • Knowledge about the relations between different models and viewpoints.
  • Knowledge of reasoning procedures to solve quantitative and qualitative problems.
  • Knowledge of first principles and knowledge to reason on their basis in order to solve novel or unfamiliar problems

Deep learning algorithms can be successfully applied to big data for knowledge discovery, knowledge application, and knowledge-based prediction. In other words, deep learning can be a powerful engine for producing actionable results.

Which was your favourite course(s) and why? 

My favourite courses were Survey Sampling and Time series because they are very useful and applicable to real life cases. My favorite part of being in a master in Statistics and Econometrics  was being challenged by professors with interesting problems, especially the real application of Time Series, and survey sampling projects.

What do you plan to do next ?

I will start my internship on April 06, 2020 in the international company IQVIA-France in Paris.

I will work as an Economic Statistician in the Real-World Solutions (RWS) Department of IQVIA France, which brings together a team of 100 multidisciplinary and highly qualified consultants in market access, real-life studies, health economics and epidemiology.  Future plan after graduation will be to find a job as a Data Scientist in Paris or Washington.

Alumni report: Joanna Morais, AVISIA

Alumni Joanna Morais

What is your position today?

I am currently working as a data scientist consultant at Avisia in Bordeaux. Avisia is a French consulting company created in 2007. Originally based in Paris, it now has agencies in Nantes, Lyon, and more recently in Bordeaux. I carry out projects ranging from a few weeks to several months, during which I accompany clients on various subjects to explain or predict phenomena, using statistical methods and data science.

I have several missions, including, for example, predicting elevator failures, the departure of customers to competitors or whether a borrower will default, determining the customers most likely to be interested in an offer, explaining the market shares of car brands according to their media investments, and creating customer segmentation according to their uses and preferences.

In parallel with the missions at the client’s site, I am involved in a number of transversal projects in my company. Those projects consist in, inter alia, participating in an image recognition project through deep learning, implementing a communication and knowledge sharing tool, organising an internal data science hackathon, and recruitment.

I also give a Marketing Econometrics course to  Master students in Statistics and Econometrics, which is a very applied course. It is an activity I particularly appreciate; it is a pleasure for me to be able to give students an idea of what concretely awaits them in the professional world,  especially for data scientist positions.

 

What was your path from your Master’s graduation to this current post, and what are the key elements which helped making your choice?

I met the company Avisia at the Business Networking Day during my first year of Master in Economics and Statistics at TSE. The following summer I did a three-month internship there; it confirmed my interest in using quantitative methods based on real data to understand customer behaviors.

I graduated in 2014 and met another company, BVA, a French market research and consulting institute. With three other students, I carried out a project with BVA for six months: the objective was to analyse the impact of media investments on the market shares of car brands.

I then completed my final internship at INBOX, another consulting firm I met at the BND. BVA came back to me afterwards to offer me the opportunity to continue the student project I had been involved in as part of a CIFRE thesis – the French denomination for a thesis financed by a company.

This thesis was a wonderful three-year experience. I was able to evolve within the company while having a research activity and very strong ties with the academic world, and in particular with TSE, since my thesis directors were Christine Thomas-Agnan and Michel Simioni. Something I also loved were the conference trips abroad – in Prague, Hong Kong, Shanghai, Chicago, Siena, … I met some very nice people, and improved my English! Moreover I won a prize for the best oral presentation at an international conference, and the Paul Sabatier prize for the best thesis in mathematics. This shows that corporate research has nothing to be ashamed of  compared to purely academic research!

I then stayed one year at BVA as a data scientist consultant. I was in charge of developing and leading an internal community of about 20 statisticians – methodological sheets, trainings, events – and the external development of the Data Intelligence offer. I also supervised a TSE student on a project to predict customer satisfaction.

Moving from a doctoral student status to being an employee is not such an easy step: I quickly wanted a new challenge. I left for Quantmetry – after being co-opted by a TSE alumni – a leading Parisian data science consulting firm, where I was able to improve my technical skills by working closely with data engineers on industrialised projects.

Some personal plans led me to leave Quantmetry and to move to Bordeaux, which coincided with Avisia’s desire to open an agency there.

 

According to your professional experience, what are the most useful skills you obtained during your degree?

Economics has given me a way of thinking about problems and understanding them. The very applied Master 2 courses with many projects – in groups or individually – the oral presentations, programming on R or Python, and professional speakers are great assets for professional integration. They prepare relatively well for the reality of the labour market. The existence of an alternative track for the M2 Statistics and Econometrics is, in my opinion, also very good.

The TSE network has been very useful to me at all stages of my professional career, whether for internships, for my thesis, or to find a job. I am now very proud to be able to maintain this link with the new generations through the course I give to M2 students.

 

What advice would you like to give to TSE students, or to the school?

My advice for students would be to really get involved in the projects they have to carry out – especially group projects – because this is by far what will best involve them in real life, and it will be an asset for job interviews.

My advice for TSE would be to maintain or even to further increase school-business relations, whether through the BND,  professional speakers, or research collaborations with companies.

 

groupe-avisia-logo