En statistique et en machine learning, on utilise souvent la notion de "faux positifs", "vrai positifs", rappel, précision ... Nous allons expliquer ces concepts aujourd'hui.
You have a chessboard of 8*8 square. You have 8 queens. Your goal is to place the 8 queens on the board, without any of them threatening another one. A queen is threaten if she is on the same row, or the same column, or the same diagonal of another queen (like in the rules of chess).
This is a project I worked on recently : detecting faces on an image thanks to machine learning (using caffe for CNN).
Today, an article about caffe, a framework for convolutional neural network (machine learning). Caffe is quite famous, but its documentation is terrific. In this tutorial we will train a model, and use the python interface to ask the Neural Network to classify an image.
Parlons un peu du passage au numérique dans l'éducation. Il y a deux-trois points qui me titillent que je voudrais partager.
A quick article to explain two new functionalities of vulgairedev.
OCCInterface and ExploREST are part of the result of my work this summer. Check it out !
The main project i am currently working on is a playground wich illustrate the possibilities of a REST API, which aim is to normalize and unify the cloud management (more infos on a next article, i guess?).
To do so, I use React. One of the features of this playground is that, like in a Documentation, you have some Readings. You click on a Menu, then you can read on a specific location of the page. In a previous project, I did it with HTML files that I included, but we wan do way better. We can use Markdown.