|
Arjun (Raj) Manrai
@
arjunmanrai
Boston, MA
|
|
Assistant Professor @HarvardMed, @Bos_CHIP. Stats+ML for improving clinical decision making and scientific reproducibility.
|
|
|
848
Tweetovi
|
268
Pratim
|
1.187
Osobe koje vas prate
|
| Tweetovi |
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Jon Gjengset
@jonhoo
|
3. velj |
|
In January, @anishathalye, @jjgort, and I ran a short class at @MIT_CSAIL on topics we think are missing in most CS programs — tools we use every day that everyone should know, like bash, git, vim, and tmux. And now the lecture notes and videos are online! missing.csail.mit.edu pic.twitter.com/xNSlLgJfd4
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
2. velj |
|
"In my opinion, the careers of Trevor Hastie and Rob Tibshirani highlight the best of what happens when statisticians interact richly with machine learning researchers." - great post on the benefits of bridging stats & ML by @IAmSamFin twitter.com/IAmSamFin/stat…
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
27. sij |
|
Excited to announce a new version of our course at Harvard, BMI704: Data Science for Medical Decision Making, and to be teaching with @chiragjp. We'll release lectures, data, and an interactive text of data science methods at link below.
Sched + Readings:
github.com/manrai/BMI704_… pic.twitter.com/fyhyZnXU75
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Isaac Kohane
@zakkohane
|
13. sij |
|
This is a patient safety issue. GC’s understand genetics and probability more than many doctors (see MD skills: @JAMAInternalMed jamanetwork.com/journals/jamai…) twitter.com/MuinJKhoury/st…
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Eric Topol
@EricTopol
|
12. sij |
|
Perhaps the least acknowledged downside of deep neural net #AI models: the carbon footprint
But this key preprint is starting to get noticed arxiv.org/abs/1906.02243 by @strubell @andrewmccallum jamanetwork.com/journals/jama/… @AndrewLBeam
technologyreview.com/s/613630/train… @techreview @_KarenHao pic.twitter.com/j3dcGityEv
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Ken Mandl
@mandl
|
10. sij |
|
So proud of @Bos_CHIP @BostonChildrens undergrad student Luke Melas Kyriazi who is now a Rhodes Scholar! rhodeshouse.ox.ac.uk/scholars/rhode… #RhodesScholarships (and great mentorship by @arjunmanrai )
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
8. sij |
|
A PAC of ML researchers? pic.twitter.com/kC5TJlBP7H
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Eric Topol
@EricTopol
|
7. sij |
|
The importance of reproducibility and replication for #AI models in #healthcare, and the under-appreciated barriers for this to be achieved jamanetwork.com/journals/jama/… @JAMA_current by @AndrewLBeam @arjunmanrai and @MarzyehGhassemi @HarvardDBMI @UofTCompSci @Bos_CHIP @VectorInst pic.twitter.com/kQam7utFEm
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Chirag Patel
@chiragjp
|
6. sij |
|
Of the 3K papers published by @NIEHS grantees in 2019, I am proud that RagGroup student @BradenTierney's work in @cellhostmicrobe was among the 26 recognized by the institute: factor.niehs.nih.gov/2020/1/feature…. Proud to be supported by @NIEHS! c/o: cell.com/cell-host-micr… cc @adkostic
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Ben Kompa
@BenKompa
|
6. sij |
|
Presenting "cui2vec" arxiv.org/abs/1804.01486 today at @PacSymBiocomp. Work done with @AndrewLBeam and @zakkohane (and many others!) on extracting medical knowledge across multiple massive data sources #PSB20 pic.twitter.com/iB6dBw0bFE
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Gary Collins 🇪🇺
@GSCollins
|
6. sij |
|
"[...] machine learning researchers moving into medical applications could [should] adhere to standard reporting guidelines such as TRIPOD, CONSORT, and SPIRIT, which are now being adapted for machine learning and artificial intelligence applications" tinyurl.com/y4kxcm8n twitter.com/MaartenvSmeden…
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
JAMA
@JAMA_current
|
6. sij |
|
At a minimum, a #machinelearning model should be reproduced, and ideally replicated, before it is deployed in a clinical setting ja.ma/35m6Aqj
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
6. sij |
|
Reproducing neural architecture search for some ML models can cost 3 R01 grants, but do most medical applications need this? Our take on the top challenges to reproducibility of machine learning in medicine in the latest @JAMA_current: pic.twitter.com/33AjE2lraY
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Maarten van Smeden
@MaartenvSmeden
|
6. sij |
|
New viewpoint worth reading about reproducible machine learning by @AndrewLBeam @arjunmanrai @MarzyehGhassemi. Reproducing machine learning models can be expensive (>1M$ just to rerun the algorithms) but vital before implemented in clinical practice
jamanetwork.com/journals/jama/… pic.twitter.com/tmdDRomowl
|
||
|
|
||
| Arjun (Raj) Manrai proslijedio/la je tweet | ||
|
Andrew Beam
@AndrewLBeam
|
6. sij |
|
New article with @arjunmanrai & @MarzyehGhassemi where we outline some specific reproducibility challenges for ML in healthcare.
Main issues:
- Data sharing
- Code documentation
- Expense of reproduction
Read the full article now in @JAMA_current: jamanetwork.com/journals/jama/… pic.twitter.com/0qYV8h3NuU
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
19. pro |
|
Incredibly privileged to work alongside this group and proud of all they accomplished in 2019. Excited for the next year! pic.twitter.com/43UgkKOwh9
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
16. pro |
|
@NAChristakis at @Bos_CHIP today demonstrating the power of human networks to shape our health and our lives. pic.twitter.com/JldCTzFaA3
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
10. pro |
|
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
2. pro |
|
Amazing opportunity for undergrads wanting to do cutting edge biomedical informatics and machine learning research at @harvardmed! twitter.com/HarvardDBMI/st…
|
||
|
|
||
|
Arjun (Raj) Manrai
@arjunmanrai
|
2. pro |
|
Honoring @MIT_CSAIL's Pete Szolovits at @harvardmed , @zakkohane shares 3 articles by Pete that were decades ahead of their time and are being recapitulated today. pic.twitter.com/7iLRjOtxoo
|
||
|
|
||