Hi! Nice to see you
I am Marcelin Seble

I am studying Data Science at Sorbonne University in Paris. I am very passionate about everything related to data processing, artificial intelligence and more. Because, I think if someone has the right data, they can learn amazing things. So I chose to learn data science in order to help leaders make the right decisions to develop the profits of their business. You can check some of my works on my github account.

Education

2020-Now
Sorbonne University
Paris, France
Master in Data Science
2019-2020
ULCO(Université du Littoral Côte d'Opale)
Calais, France
Bachelor degree in Computer Science

Experiences

Mar. 2022 - Aug 2022
Developer Analyst intern
  • Design a pipeline for extracting, processing and exploring large raw files containing billing data.
  • Create a dashboard to produce relevant indicators to analyze customer consumption on many dimensions such as time, products, countries, etc.
  • Attending in meetings on technology intelligence
Feb 2019 - Aug 2019
Full stack developper in .Net
  • Development of a secure multiplatform application (web and mobile version) for inventory and order management.
  • Contribution to the design of a VAT tracking system allowing Togo's tax authorities to follow the VAT information related to taxpayer sales. I have much more work on the data integration and data visualization part via a web application.
Apr 2018 - Nov 2018
Business Intelligence intern
  • Extraction and processing of transaction data phone numbers from Moov Togo subscribers.
  • Data integration in Data Warehouse
  • Creation of an OLAP cube allowing to study the transactions(calls made and received, SMS, internet package purchase, unit credit purchase) by date, location, phone number and so on...
  • Automation of extraction, processing and data integration(the objective was to synchronize the new arrival files from the retrieved information by relay antennas).
  • Creation of a web application for monitoring and generating insights for administrative meetings.

Works

Autoencoder from scratch

Autoencoder from scratch

Design Autoencoder from scratch. The goal of this project is to design from scratch a neural network layers(linear or convolutional (1D)) and activation functions to realize an auto-encoder.

More

Search engine using CACM collection

Academic project about information retrieval and the goal is to build a search engine using CACM data collection to train our models. We did indexation, create models(Okapi, Vector model), make evaluation and implement page rank algorithm.

Search engine using CACM collection
Business Intelligence project

Business Intelligence project

For this project, our goal is to collect data in order to find an answer for "What is the most suitable US state for raising children?". To do it we use Talend and Dataiku.

More

SAT Solver(CLI Version)

This is a small solver based on DPLL iterative algorithm. You can install(cli version) this solver and use it to figure out propositional satisfiability problem (SAT). For example, it can help you to establish championship planning for each team according to some constraints.

More
SAT Solver(CLI Version)
ML Projects

ML Projects

All projects about machine learning I do is here.

More

Contact

Email

marcelinseble@gmail.com

Address

Ile-de-France