I'm Ahmet Akkoç, MSc Data Science grudate from the IT University of Copenhagen 🇩🇰.
More about me:
* I love #tech, #wayofthefuture
* #opensource and #Fediverse enthusiast, but cautionary of #Mastodon :3
* #Linguistics enthusiast, here to promote it on Mastodon.
* I get around, expect to see me all over the place!
Pleasure to meet you (*^_^)/
| My Homepage (Academic) | https://ahmet.akkoc.online |
| My Other Homepage (Software) | https://madpro.neocities.org/ |
👀Looking into the future! 📈 Just wanted to share our patient outcome prediction model which we presented at EULAR 2025.
We have designed a #forecasting #model that can follow a patient's swollen joint count over time and estimate if their
condition will improve or worsen.
ZiteLab #pytorch #sktime #opensource #personalizedmedicine #ai #ml #timeseries #rheumatology
(1/2)
1/2 Credit where credit is due? Not so much in the world of #ai !
In our latest paper for #MIDL , we examine how many researchers fail to cite datasets which they use.
Medical imaging papers often focus on methodology, but the quality of the algorithms and the validity of the conclusions are highly dependent on the datasets used. As creating datasets requires a lot of effort, researchers often use publicly available datasets, there is however no adopted standard for citing the datasets used in scientific papers, leading to difficulty in tracking dataset usage. In this work, we present two open-source tools we created that could help with the detection of dataset usage, a pipeline \url{https://github.com/TheoSourget/Public_Medical_Datasets_References} using OpenAlex and full-text analysis, and a PDF annotation software \url{https://github.com/TheoSourget/pdf_annotator} used in our study to manually label the presence of datasets. We applied both tools on a study of the usage of 20 publicly available medical datasets in papers from MICCAI and MIDL. We compute the proportion and the evolution between 2013 and 2023 of 3 types of presence in a paper: cited, mentioned in the full text, cited and mentioned. Our findings demonstrate the concentration of the usage of a limited set of datasets. We also highlight different citing practices, making the automation of tracking difficult.
Presenting #prognosisAI at the Scandinavian Congress of Rheumatology this week.
Some photos from today's #HealthDataScienceDay
conference in #copenhagen. I presented my thesis on differential #privacy and how the #algorithm design impacts whether trends for subpopulations are preserved in resulting synthetic data.
Read more here: https://bit.ly/3NovQnK
My Poster for tomorrow's #HealthDataScienceDay at Mærsk Tower, #Copenhagen. It's all about #differentialprivacy !
@UCPH_health
I keep hearing people talk about how great #gpt3 is at multi-lingual translation.
But the matter of fact is that it's not designed to be consciously multilingual. When given a rare language like Zhuang, it will delude itself and you as the user into thinking that it is made-up.
What is interesting, is that given a Hanzi transliteration of the language, it can translate it assuming it to be Chinese. #LogographMyLanguage
Sneak peak from my #privacy research:
A network of people, adblockers and where they find them.
I am mapping the relationships between #consumers, #privacy goods and services and suppliers.
Can you read the story being told in these figures? How would you tell it?
I recently learned that in #APA, it is now possible to reference websites without titles and even social media posts on Instagram, Twitter and Facebook:
https://libguides.usc.edu/APA7th/website
https://libguides.usc.edu/APA7th/socialmedia
More awkward #ChatGPT output.
Again, items in sequences are handled in a very formulaic way:
"X is a Y newspaper that generally supports Z causes and has been critical of T."