Ceder Dens

[Ceder Dens @CederD]

Beware of #Bias in training data significantly impacting real-world #MachineLearning performance! Our study in @NatMachIntell reveals how bias in negative data generation for #TCR - #Epitope binding prediction causes #ShortcutLearning
https://t.co/YgAz12i3zf

[Pieter Meysman @chevaliersf]

Our new paper on deciphering how complex #DeepLearning models make #TCR - #epitope predictions is out on biorxiv! Great job by Ceder Dens! #AIRR
https://t.co/43WOw2trRo