Our new paper is out in @EGU_GMD ! We developed NN-TOC v1, a deep learning model for global prediction of total organic carbon in marine sediments. Check it out: https://t.co/SU5pATgN1E #MarineSediments#DeepLearning
Just as I learnt over the last 3 years of my PhD, deep learning models struggle with tabular data compared to tree based models. But it seems like a time for deep learning models to give a comeback. Tabular foundation models like TabPFN seem to be the game changer now!
This might be the first time after 10 years that boosted trees are not the best default choice when working with data in tables.
Instead a pre-trained neural network is, the new TabPFN, as we just published in Nature π
What can be more annoying? You have a zoom meeting. You are late and have a minute to join. And then it won't open unless you update and it takes a million year to get updated. #ZOOMEX@Zoom#phdlife
@unraveaero Land reclamation is always a loss of marine diversity and change of coastal dynamics. Singapore has land constraints which is not the case with Mumbai. With enough decentralisation, we can shift the population to satellite towns or cities on the Delhi Mumbai corridor.
Is it normal to have an imposter syndrome, when you write a paper, citing works that you feel you could never achieve, doubting your own results if it is good enough? #PhD#PhDvoice
An old podcast and a very famous paper from a PhD student! Definitely worth listening to and motivating! Great talk by @jefrankle on his PhD.
https://t.co/oHR2eEr9xi
First #EGU23! Had a lot of valuable inputs and was surprised to see so many interested people in my work!:)) Still overwhelmed by the amount of research in geosciences! @data_mar @GEOMAR_de