ICANN 2021 Replay: dive into data imputation with award-winning DAEMA
Dealing with missing values in tabular datasets is a classic challenge for data scientists and machine learning practitioners. But how do you properly and accurately impute these missing values to ensure your models perform at their absolute best?
If you missed it live, we are thrilled to share the conference replay of Simon’s talk at ICANN 2021, where he dives deep into this exact problem!
In this session, Simon presents our team's research on advanced data imputation techniques, specifically walking through the mechanics of DAEMA (Denoising Autoencoder with Mask Attention). The talk breaks down how to effectively tackle missing data in tabular formats and elevate your data preprocessing pipelines.
An Award-Winning Contribution
We are incredibly proud to highlight that the underlying research paper was recognized as the 3rd Best Paper out of 265 accepted submissions at the conference!
A massive congratulations to the team behind this work: Simon, Usama, Nicolas, Damien, and Thomas.
Dive into the Research
Want to explore the methodology and findings in detail? The full paper is freely accessible in its preprint form here: https://arxiv.org/abs/2106.16057
Watch the Replay
Tune in to the replay, read the preprint, and let us know your thoughts on the future of data imputation!