“Catherine” Chunyang Li

Statistician, Data Scientist, ML/DL Researcher, Research Associate @ University of Utah Health

[“Catherine” Chunyang Li @chunyangli_cli]

Presented at MLHC in Toronto on our deep learning survival analysis work in non-metastatic castration resistant prostate cancer patients. Thanks for the great opportunities to communicate with many people. Looking forward to participate again in the future!

[“Catherine” Chunyang Li @chunyangli_cli]

Please let me know if you think there are any other factors I didn't take into account.. #tax Here is the GitHub link: https://t.co/ashMGz0fiL

[“Catherine” Chunyang Li @chunyangli_cli]

I wrote an R function to recommend if you should choose to do Roth IRA or traditional IRA for those who want to figure this out quantitively. Github repo attached. (Perhaps an excel formula or a shiny app might be more useful for general public.)

[“Catherine” Chunyang Li @chunyangli_cli]

People's blog posts mostly just explain in a qualitative manner. I don't know if I am missing anything, my conclusion is that: it seems like for younger investors, Roth is a better option; for people who are closer to withdraw the funding, traditional IRA is a better deal...

[“Catherine” Chunyang Li @chunyangli_cli]

@stephaniehicks If in the future it becomes the trend of stay-home husbands to take care of kids, and in charge of all the house work, perhaps the graph would be a little more balanced. 😄

[“Catherine” Chunyang Li @chunyangli_cli]

It would be nice that the legal system is designed where AI judges should assist human judge in all legal matters. As described in "Talking to Strangers" by Malcolm Gladwell, AI judge shows less biases than human judge in some tested legal cases. #Legal

[“Catherine” Chunyang Li @chunyangli_cli]

@fchollet Programming is just a language. The spirit one expresses with language could be more important, that is algorithm. A language without spirit could not accomplish much. Making kids learn more math rather than programming is definitely more helpful.

[“Catherine” Chunyang Li @chunyangli_cli]

@HeidiBaya How about just simply save the trained model and then later reuse it? Since there is so much work to make the whole process non-deterministic...

[“Catherine” Chunyang Li @chunyangli_cli]

Nice comprehensive tips on: How To Improve Deep Learning Performance https://t.co/68QJsXJnxD

[“Catherine” Chunyang Li @chunyangli_cli]

Theory of Saturated Neural Networks https://t.co/YKVVKP6Lpn

[“Catherine” Chunyang Li @chunyangli_cli]

Happy happy birthday! @ahalvani

[“Catherine” Chunyang Li @chunyangli_cli]

I am listening to the book “mind unraveled” when it talks about how our experiences change us and made us a different person. All in my mind is that this book is so much related to a recurrent neural network 😅

[“Catherine” Chunyang Li @chunyangli_cli]

I dedicate this nerdy weird tweet to @ahalvani for his birthday, the person who get me so obsessed with DL that I now am reshaping my world with DL pipelines.😅May your RNN cell this year generate any outcome you desire. And countless great RNN cells to come. (Live forever😄)

[“Catherine” Chunyang Li @chunyangli_cli]

@fchollet Good applications might be more meaningful than most of technique novelty. Most of technique novelty might not be very valuable, but some might be extremely valuable. Although the scientific breakthroughs must be standing on the shoulder of the many "useless" technique novelty.

[“Catherine” Chunyang Li @chunyangli_cli]

Introduction to Markov chains by @roccajo in @TDataScience https://t.co/7bcx2vdCAo

[“Catherine” Chunyang Li @chunyangli_cli]

Interpretable Machine Learning https://t.co/lA3mm8QZZY #rmarkdown #bookdown