data from the trenches

We introduce the MQ Forecaster & Transformer models and compare their performances against classic ML methods.

Using RNNs & DeepAR Models to Find Out

This is the story of a research project that didn’t quite make it. We introduce a new active learning strategy and put it to the test.

For robust causal estimation, one cannot blindly use causal model. We present tests to check plausibility of the main causality hypothesis.

To best plan a personalised marketing campaign, we’ll see how Causal Inference can help tackle churn retention.

Scheduling at ALMA relies on accurate weather forecasts, can we apply Deep Learning to fine tune the forecasts in the Atacama desert?

Combining ML and drones, a data scientist dream. We’ll show how to deploy a SSD model on a drone with Uavia DroneOS and Dataiku DSS.

Your data has drifted, and you now need to determine the source of the drift. This can be achieved leveraging interpretability techniques.

Augmenting any model with an uncertainty estimation? It might sound too good to be true, but that’s the promise of conformal prediction.

We give an in-depth overview of the various definition of fairness for Machine Learning. Spoiler alert: we can’t have them all.

Deep learning models are known for their impressive size, but how do you deploy those on edge devices where size matters?

Detecting drift is critical in the monitoring of deployed models. So, how do we efficiently detect data drift ?