Light DJOSSOU

AI & ML Engineering

[Light DJOSSOU @light_ai667]

Semaine 5 de mon Experience Lab chez @analystlabafric : de la conception (Week 4) au code qui tourne, sur le projet HealthConnect.

Ma piste : ML Engineering. Objectif : implémenter le pipeline conçu la semaine dernière.

#AnalystLabAfrica

[Light DJOSSOU @light_ai667]

🏥 Week 4 of my Experience Lab @analystlabafric
New shared project (HealthConnect) on my ML Engineering track.

Found a data leak before writing any code: “waiting time” was filled in for patients who never showed up 👀

System architecture delivered ✅
#AnalystLabAfrica #ML #AI

[Light DJOSSOU @light_ai667]

Semaine 3 de mon ML internship chez @analystlabafric !
J'ai entraîné et comparé 5 modèles pour prédire le churn client (Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, XGBoost).
Random Forest recommandé : ROC-AUC 0.836, Recall 74% 🏆

#AnalystLabAfrica #ML

[Light DJOSSOU @light_ai667]

🤖 Week 2 of my ML internship at @analystlabafric
I cleaned a dataset of 7,043 Telco customers... and discovered 11 totally invisible missing values 👀
Feature engineering + encoding + outliers + feature selection ✅
100% ML-ready dataset 🚀
#AnalystLabAfrica #ML #AI #DataScienc

[Light DJOSSOU @light_ai667]

Week 1 of my ML internship at @analystlabafric

26.5% churn rate at a (fictional) telecom operator. But the real insight? 11 hidden missing values in a dataset that showed "0 NaN"
Business framing + EDA + proposal → all up on my GitHub : https://t.co/YvBqI6pLcp
#ML #AnalystLab

[Dima Zeniuk @DimaZeniuk]

Starships at Starbase

[Tesla Europe, Middle East & Africa @teslaeurope]

Better in all the ways you were hoping for

New Model Y on display across Europe starting tomorrow – come have a look in person

[SpaceX @SpaceX]

Launch and return are fundamental techniques for Starship’s fully and rapidly reusable design