13th meeting

• subject #1 : Deploy machine learning model into web app. Speaker : Peter HIRT

• subject #2 :

Interactive topic modeling with LDA. Speaker : Quentin PLEPLE

• excerpt to subject #1: On the example of a (relatively simple) ML algorithm for House Price predictions, the methods on how to use the model in a web app to predict house prices is shown. The ML part is done in Python, the model is deployed to Amazon Web Services on a S3 instance while the actual webapp is done in AngularJS on the FE and a on Apache on the BE.

• excert to subject #2:

Topics discovered by the latent Dirichlet allocation (LDA) method are sometimes not meaningful for humans. To improve the quality of topics presented to end-users, our novel method allows the user to give live feedback on the topics, and allows the inference algorithm to use that feedback to guide the LDA parameter search