About me

pwd

/home/artorithme/myWebsite/

whoami Jean-Pierre Axel Gninahophin

cat activities.md I spend most of my time messing with my computer when I’m not biking.

chmod +x presentationScript.sh ./presentationScript.sh WELCOME TO MY BLOG !

You’re currently on the about me page!
Click on the Artorithme logo to explore my blog!

What I’m good at

coding

Programming

This is what I do during my weekends
When I want to have fun
If I’m a not programming, I’m probably reading/writing about it.

Check out my blog!

coding

Analysing Data and explaining Models

I really enjoy exploring data, and sharing insights I gain from it.
I like building fancy models, using Neural networks and Xgboost, but what I like the most is making them understandable to drive better decisions.

Education

2020

PCSI-PSI

Introductory class in Mathematics, Physics,Engineering


  • Analysis, Algebra, Geometry, and Probability
  • Mechanics,Thermodynamics,Electricity, Magnetism,Fluid Mechanics and Optics
  • Computer Science (Python) and Mechanics

2022

Actuarial science - DUAS

Advanced classes in Probability Theory, Statistical Analysis, and Risk Modelling
Introductory classes in Finance,Insurance and Investments


Machine Learning Engineer Certification

Codecademy

2024

Master in Signal and Image Processing : Specialization in Computer-vision






Programming languages I can’t leave without

R

R is my favorite language by far and the main reason I created this blog. I think people don’t realise how fantastic it is because they discover it in Academia with a statistic teacher who isn’t really interested in the programming aspect.

Javascript

Javascript is a mess. Everyday,there’s a new UI library or a framework… Npm will have you dowload the whole internet before you can center a div. But that’s the price to pay to have nice things on the internet.

Python

Python doesn’t really need any presentation.

Bash

I’m a Gnu/Linux user and no, I don’t use Arch (by the way) because I have matured and I use Debian as all my sysadmins friends. For me, Bash is a lifestyle.






A 2D array of my latest endeavors

Title Tech Tech stack Details
My personal website HTML, CSS, Javascript,Bash,Python Jquery, Bootsrap,Pandoc,pandas
Monte-carlo Simulator C++ STD A Monte Carlo simulator in C++ designed to estimate the expected loss or gain from investing in a security whose value varies according to a leptokurtic distribution.
Advanced social media Web-scrapper Python, Javascript, Chromium Selenium, Pandas A web scraper that mimics typical user behavior to gather data and assist in combating scam artists.
Reverse engineering an api to scrape chess games and train a Chess-bot Bash, Php jq, curl
Insurance pricing dashboard HTML, CSS, Javascript, R shiny, data.table, Bootsrap I significantly optimized the performance of a complex pricing tool by migrating it to R, resulting in real-time update capabilities, and developed an intuitive R dashboard to simplify user interaction and enhance overall all within a remarkably short timeframe.
Recovered a non-booting computer by resolving a kernel failure and repurposed it as a functional Linux backup server. Bash, OpenSSH
Developed and optimized a multi-layer neural network model to predict wine quality, utilizing the R port of PyTorch, and the H2O library R, Python Torch, Pytorch, R6, H2O, data.table Trained a multi-layer neural network on a severely imbalanced dataset, leveraging insights from relevant literature to develop a high-bias, low-variance model that effectively distinguished high-quality wines from the rest, despite the datasets inherent class imbalance.
Designed a modular framework leveraging R6 object-oriented programming and grid search to programmatically generate and evaluate multiple model configurations, enabling the identification of the optimal model architecture. R, Python Torch, Pytorch, R6, H2O, data.table To support my research, I created a tool to programmatically generate neural network models in Torch, which enabled me to systematically explore different architectures and identify the optimal configuration. Through a grid search on a subspace of multi-layer neural networks, I was able to find a model that effectively balanced complexity, performance, and prediction time.
Conducted an in-depth analysis of NFC technology NFC, C, Bash Delved into the inner workings of NFC technology, examining its fundamental physics principles and circuit design to understand its behavior. Furthermore, conducted a security-focused technological watch to evaluate the resilience of NFC-enabled devices against potential threats, simulating attack scenarios and identifying areas of vulnerability that could be targeted by attackers.
Conducted a data-driven analysis to investigate the relationship between smoking habits and social category in an insurance portfolio, challenging common heuristics that could lead to unfair pricing practices. Python, R, scikit-learn, data.table, matplotlib, seaborn
Developed a photoplethysmography algorithm utilizing computer-vision techniques to extract blood frequency signals from smartphone video recordings Python, OpenCV
Designed and implemented data engineering solutions for financial data related to the European Union Emissions Trading System (EU ETS) Python, R data.table, pandas, scrappy, requests Conducted a comprehensive review of existing literature on the European Union Emissions Trading System (EU-ETS), and collected, processed, and analyzed relevant data to support further research and analysis during an R&D internship.
Image Compression using Singular Value Decomposition (SVD) Python numpy