Olatomiwa Bifarin, Ph.D.

Biochem PhD, Postdoc @georgiatech.
I work in the field of #metabolomics using #machinelearning and #datascience techniques.
ML blog:

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

I recently gave a talk at @GeorgiaTech College of Sciences data science seminar.

My talk was titled "Machine learning Applications in Metabolic Phenotyping." Available to watch here. https://t.co/ELx5zyro2r

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

My next series of blog on The Epsilon will be on this architecture.

As such, in this blog, I write about the operation at the heart of Transformers, Self Attention.
https://t.co/PQGOaeQG5F

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

Most people, even those living under the proverbial rock, have used apps like chatGPT.

But why are these technology so powerful?

Short answer: they use a powerful neural network architecture called Transformers.

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

My latest AI newsletter covers publications in SC foundation models, Google's new multi-modal AI model, ideas about LLM OS, amongst other.

ε Pulse, Issue No 8: Gemini 🤖, GenePT 🧬, and Rebuilding Organization with AI 🏢.

https://t.co/q42zP3letq

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

In my last blog, I wrote about the theoretical underpinnings of SHapley Additive exPlanations (SHAP). In this blog, we will see to a couple applications: interpreting a SVM and a language model.

Blog: https://t.co/xBmUVkQ68c
Supplemental code: https://t.co/VGyFNf9caG

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

@BiswapriyaMisra @facundofGT Thank you!

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

two recent preprints from my postdoctoral research:

1) I developed a pipeline that harmonizes AutoML technique with the interpretations of resulting ensemble models in metabolomics data science: https://t.co/FHVhWxsyTm

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

In my latest entry on The Epsilon, I continue my series on explainable #AI.

I write about Shapley values and Shapley Additive Explanations.

Hope you enjoy reading.

[https://t.co/3hGr49aC8s]

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

In the next couple of blogs, I will explore some explainable AI techniques.

I am starting with LIME largely because it will be a nice segue into the more popular technique, Shapley Additive Explanation (SHAP).

[https://t.co/WQDpOycnLl]

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

Latest issue of my AI/ML newsletter:
[https://t.co/YDPhYul0Fr]

[Bif @BifarinTheFifth]

Latest issue of ε Pulse:

[https://t.co/8SRy1FZA14]

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

15/n:
There are many, many more research I find interesting, but better keep the thread short and sweet(?)

It’s being an amazing #ASMS2023!

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

1/n Lots of interesting talks/poster in #ASMS2023, here are some of my favorite, so far. (FYI: informatics-focused)

A 🧵

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

1/n Lots of interesting talks/poster in #ASMS2023, here are some of my favorite, so far. (FYI: informatics-focused)

A 🧵

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

14/n: Eight: Automated ML model selection in metabolomics plus model agnostic interpretations with kernel SHAP; applied to a kidney cancer urine-based LC-MS dataset. #ASMS2023
👇🏿
https://t.co/jss7a4zugS

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

If you are interested in automated and interpretable machine learning for MS Metabolomics, come by my poster tomorrow and let’s chat! ThP 326 #ASMS2023

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

Seems to hit the mark! #ASMS2023

[J. of the American Society for Mass Spectrometry @J_ASMS]

#ASMS2023 🫶🏻

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

#ASMS2023

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

#ASMS2023

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

#ASMS2023

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

If you are interested in automated and interpretable machine learning for MS Metabolomics, come by my poster tomorrow and let’s chat! ThP 326 #ASMS2023

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

Latest work from my postdoctoral training with @facundofGT at @GTSciences

[Journal of Proteome Research @JProteomeRes]

🔓 #OpenAccess: "Machine Learning Reveals Lipidome Remodeling Dynamics in a Mouse Model of Ovarian Cancer"
https://t.co/WhnK9ZwbPK

[Olatomiwa Bifarin, Ph.D. @BifarinPhD]

@SalmaAboElhasan Also, I wanted to point you to our latest paper: the computational pipeline might help you analyze your data much better: https://t.co/oQHsG5zICk