Overview

This repo contains code for supervised classification based on text files and labellised pictures Pictures labellisation is run with AWS REKO API Model is trained with AUTOML

steps

1. get text datas (csv) and associate with pict to build training set : 1-import-datas.ipynb
2. labellise pict (batches of 100 pict) : 2-AWS_REKO.ipynb
3. concat labels with training set : 3-concat-datas-with-AWS-labels.ipynb
4. train model and predict target : 4-predict-groupe-AutoML.ipynb