I'm hiring Research Scientists in DeepMind to work on the intersection of ML, science and healthcare. Apply for the
@DeepMind
RS role and express interest to work with me here: . My DMs are open, so feel free to reach out to me if you have questions.
An update! Two weeks ago I started my journey as Vice President of AI/ML at GSK AI. I’m really excited by this opportunity to collaborate with an amazing team, focusing on leveraging AI to improve patient outcomes.
Thanks to everyone who submitted tutorial proposals for this year's
@NeurIPSConf
. It was amazing to see all the great work that went in to putting together these proposals. We are happy to announce this year's tutorials at
#neurips2020
Understanding how we can use human-centric machine learning for improving mental health is something that I’m really passionate about. Excited to share our latest work on using ML to understand engagement with cognitive behavioral therapy
@JAMANetworkOpen
Thanks to everyone who attended our
#ICML2018
tutorial on Machine Learning for Personalised Health given by
@konstantina_pll
, Lamiae Azizi and myself. A copy of our slides can be found on the tutorial website here:
And it's a wrap! My best
#NeurIPS2022
so far. So grateful for the opportunity to work with and learn from my amazing program co-chairs
@aliceoh
@kchonyc
and Alekh. It's been great to have in person discussions on all things ML. Looking forward to the virtual sessions next week!
Are you attending
#NeurIPS2023
?💡
I am actively recruiting for a number of open roles at different levels in my team at
If you are interested in leveraging AI to understand and advance solutions in health 🏥 please reach out! ✉️
Here's a copy of the slides from my presentation
@black_in_ai
workshop during
#NIPS2017
"Machine Learning for Personalised Health": . Thanks to
@timnitGebru
@red_abebe
and organisers for putting together such an amazing workshop.
Reflecting on
@katecrawford
's
#NIPS2017
tips for fairness in AI:
1. Is the dataset demographic representative of the general population
2. Take interdisciplinarity seriously and create spaces for collaboration
3. We need to think deeply about the ethics of classification
Looking forward to the
@DeepMind
booth at
#ICML2022
where we will be presenting the Deep Learning Health Portfolio at 3pm-5pm ET today and 10am-12 noon on Wednesday. Come for a chat about all things
#ML4Healthcare
and meet the team!
280 characters are insufficient to express how excited we are that
@WiMLworkshop
!starts in 1 week!!!!! We have an amazing line up of speakers, contributed talks and mentors with opening by WiML Co-Founder
@jennwvaughan
. Check out the program on:
#WiML2017
Looking forward to great tutorial proposals for
#NeurIPS2020
(the call for proposals will be out soon!). Bob Williamson and I (Tutorial co-chairs) want to know, what topics would you like to see at this year’s tutorials?
I'm excited to share some of the work we've been recently doing at at the intersection of AI and precision oncology, accelerating the biopharma pipeline, and agent-based operating systems
We have a fantastic line-up of tutorials
@NeurIPSConf
2020 tomorrow! As well as the Q&A and presentations on Monday, check out the 1 hour Q&A sessions on Tues, Wed and Thurs where you can also ask tutorial presenters questions on their tutorials.
One of my favourite highlights of
@NeurIPSConf
is
@black_in_ai
workshop. I remember being overwhelmed by this great community and allyship when I gave a keynote at the 1st workshop in 2017. I'm looking forward to running a mentoring table with
@jennwvaughan
tomorrow at 2:30pm EDT
Here is the schedule for the 4th
@black_in_ai
workshop happening this Monday, December 7th 2020!
We have amazing speakers from the black diaspora ranging from Brazil, Nigeria, UK, USA who will cover topics including theoretical AI advances, startups and social science
We are happy to announce the
@NeurIPSConf
#NeurIPS2020
Call for Tutorial proposals Bob Williamson and I (tutorial chairs) have written a blog post summarising changes to this year's tutorials and tips for writing a great proposal
Such amazing work presented by Claire Monteleoni on Machine learning for
#ClimateChange
. Prof Monteleoni is one of the trailblazers of this field and started the international workshop of climate infornatics in 2011. We're so lucky to have her here
@Khipu_AI
It was great to be part of the
@ghdatascience
#indabaxghana
where I talked about building successful pipelines for predictive analytics in healthcare. It was great to speak to the participants and see the innovation, creativity and energy in the AI space in Ghana.
Predictive Analytics in Healthcare:
Design is one of the most important aspect of a study and perhaps the most neglected.... the study design should match the research question - Danielle Belgrave, PhD
#indabaxghana
#ghdatascience
We are looking for mentors to volunteer for
@Aistats2020
. If you would like to host a 1 hour zoom mentoring session, email us at aistats2020dfi
@gmail
.com
Deep reflection by Cynthia Rudin on recent algorithms on for interpretable ML at
#icml2018
Fairness, interpretability and explanability workshop
- The definition of interpretability is context specific
- it's harder to get an interpretable model than to get an accurate model
@shakir_za
spreading his energy to the
@ghdatascience
beforw his talk on Statistical ML and applications. It's great to see the
#indabaxghana
in the context of building the data science and ML community together on the African continent.
Check out our new website for
@MSFTResearchCam
Healthcare Intelligence team It's great to be part of such an amazing team. If you are passionate about machine learning for healthcare using a human-centred approach, we're hiring!
Really looking forward to seeing this journey unfold! This is super exciting news for the future of drug dicovery and medical and biological research using an AI-first approach.
Thrilled to announce the launch of a new Alphabet company
@IsomorphicLabs
. Our mission is to reimagine the drug discovery process from first principles with an AI-first approach, to accelerate biomedical breakthroughs and find cures for diseases. Details:
Thought provoking keynote talk by
@CwaMaina
at
#DALI2019
in South Africa on initiatives to increase the use of
#MachineLearning
to solve problems within the local context. Important lessons from Data Science Africa as a grassroots initiative.
@zacharylipton
Mathematical statistics and data analysis by John Rice. Intro to Mathematical Statistics by Hogg & Craig is also good, but the proofs aren't as detailed as in Rice.
We are happy to announce the
@NeurIPSConf
#NeurIPS2020
Call for Tutorial proposals Bob Williamson and I (tutorial chairs) have written a blog post summarising changes to this year's tutorials and tips for writing a great proposal
So amazing to see the work
@black_in_ai
are doing to solve problems on the African continent.
@GeorgeWMusumba
speaking on how machine learning can be used to help improve outcomes in the construction company. He gives an excellent end to end evaluation of the problems involved.
One of my favourite parts of the
@WiMLworkshop
today was the mentoring round tables discussing research topics, career advice and
#NIPS2017
papers. Thanks to all the amazing mentors at
#WiML2017
who gave their time and energy!
I really enjoyed presenting at the Data Science Summer School
@Polytechnique
on advances and challenges in machine learning for healthcare Thank you
#xDS3
- for such great community and conversations on ML and healthcare. Slides from my talk are here: )
The Women in Machine Learning Workshop 2018 Organizer Application is now live! Deadline-25th March. I had a lot of fun co-organising this last year and I got to work with a bunch of amazing women like
@negar_rz
and
@_ehinosa
.
#WiML2017
#WiML2018
#NIPS2018
Interesting dialogue between
@katecrawford
and
@hannawallach
on
#fairness
at
#ICML2018
. If we use data from a system which is structurally biased, we may end up optimising an unjust system. We need interdisciplinary ML research to understand ways in which data can be skewed.
Really looking forward to speaking at the Advances in Data Science meeting in Manchester in May and exploring how data science can support societal wellbeing.
Happy to announce our
@icmlconf
2020 workshop on Machine Learning for Global Health. If you are working in global health, please consider putting forward an abstract. We have an amazing line up of speakers working on challenges around global health:
Brilliant talk by
@AndrewYNg
at the
@mlforhc
conference. Here is a summary of his talk. He has the art of being an engaging speaker - with a notebook in his hand and an iPad for a white board. He talls about not just doing research in health, but deploying to practical settings
Joelle Pineau sharing important lessons on reinforcement learning for understanding adaptive interventions for improving the quality of healthcare
@WiMLworkshop
#WiML2017
#NIPS2017
Really looking forward to giving an invited talj on machine learning for healthcare
@strangeloop_stl
today. Looking forward even more to hearing
@simonpj0
@MSFTResearchCam
speak about shaping children's education through computing.
Loving the opening remarks by
@BrownSarahM
@neurips
workshop on navigating the broader impacts of AI research. She highlights the importance of training for
#AI
researchers in developing a universal language for thinking of the broader ethical implications of their work.
Nyalleng
@nunuska
wrap up
#AIforsocialgood
panel. She says that we need to take account of skills of local communities in solving local problems, even if those solutions are not AI. Also problems in local communities are not "their" problems, but these are "our" problems
Someone in the crowd (not sure the name) has made the point (I think!!) that when you embed yourself in the real world, when you work closely with the people you "want to help", solving the right problems becomes easy, because you're solving your *own* problems
#NeurIPS2019
Applications for internships in Machine Learning for Healthcare at Microsoft Research Cambridge
@MSFTResearchCam
are now live! A fantastic opportunity to be part of a great research environment.
The Dagstuhl Declaration on
#AIforSocialGood
is now live. It's so important that
#ML
,
#AI
and NGOs form deep collaborations for
#SocialGood
. Really proud to be a part of this and to dialogue with leaders on the intersection of these fields.
Really interesting talk by Ida Momennejad
@criticalneuro
on her great work on predictive cognitive maps with multi-scale successor representations and replay in the Bioligical and Artificial Reinforcement Learning workshop at
#Neurips2019
Great to hear the voices of domain experts on Navigating the Broader Impacts of AI Research at
#NeurIPS2020
. Hearing these voices during
@saiphcita
keynote talk on the future of work for the invisible workers in AI has been one of the highlights of
@NeurIPS
for me
This 100x over. I remember the 1st time on the
#NeurIPS
organising committee as tutorial chair with
@aliceoh
in 2019. The highlight for me was getting to work with and meet amazing people while contributing to shaping the field. The rewards are high and 100% worth the effort.
Organizing a conference the scale of
#NeurIPS2024
will take a lot of work. But you get to work with really great GCs
@LesterMackey
and Amir Globerson, as well as SPC
@DaniCMBelg
@lgmoneda
@Khipu_AI
@yudapearl
@eliasbareinboim
@_MiguelHernan
So great to see enthusiasm about using DAGs for missing data and causality! The reference for this work is here: This is excellent work from Dr Alexina Mason and Profs Nicky Best and Sylvia Richardson. Excellent and systematic approach
Powerful talk by
@Abebab
on "Algorithmic Colonization" at
#Neurips2019
#ML4D
.
-Sometimes blindly applying AI technologies for social good > alluring than confronting ugly social realities
-
#AI4SocialGood
should consider of local customs and norma
We need to hear more of this.
Golden words by
@katecrawford
on Bias in AI
#NIPS2017
The perpetuation of stereotypes is harmful. We therefore need to think of the role of machine learning in perpetuating harmful representations.
Excellent talk by
@CsabaSzepesvari
at the
#CausalML
#ICML2018
workshop on causality from a reinforcement learning perspective. Interesting perspective of his understanding
@yudapearl
statements on ML
and the mutual learnings possible between RL and causality.
It was fantastic to be part of the conversation
@mlforhc
this year. Thanks to the organisers for an amazing conference and for the invitation to give a keynote.
#ml4health
is a great community to be a part of
Absolutely loved this keynote by Zak Kohane at
#NIPS2017
#ML4H
workshop. Favourite moments:
- Thoughts around the difference between an AlphaZero system and understanding health
- Machine learning in health is not just about prediction but about decisions that impact human life
Looking forward to this event tomorrow as part of the broader conversation with
@anthtechconf
on responsible AI in the context of design and development of machine learning for mental health and healthcare more generally
Join us online Weds 9 Sept at 18:30 BST to hear
@DaniCMBelg
talk about Human-Centred Machine Learning in Healthcare. We sold out so we’ve added another 25 free tickets. Grab yours quick!
@JCornebise
Understanding how patients interact w/ digital mental health interventions could mean more personalized treatment. Together w/
@SilverCloudH
, researchers built a probabilistic graphical model to ID engagement subtypes & their predicted clinical outcomes:
A powerhouse of speakers at the
#NeurIPS2019
workshop on “Do the right thing”: machine learning and causal inference for improved decision making Listening to Susan Murphy Challenges in Developing Learning Algorithms to Personalize Treatment in Real Time
Looking forward to
#WiML2017
mentorship tables on Monday with an amazing line up of mentors. Have you decided which sessions you will attend? Check them out at
@WiMLworkshop
#newlowerbound
#280
<K(WiMLexcitement)
Great to see important conversations about how AI impacts civil liberties happening at the
#AAAI18
#AIES
conference. 👏🏼 to Carol Rose of
@ACLU_Mass
for a stellar keynote!
Amazing talk by Ali Rahimi "It's weird when the things you are trying to approximate work worse than your approximation"
"Machine learning has become alchemy. [But we use ML for health]. I would like to live in a works where systems are based on scientific rigour "
#NIPS2017
Nice, classic summary of pros and cons of Bayesian Learning by
@yeewhye
giving the Breiman Lecture at
#NIPS2017
. Proposing methods of Bayesian Deep Learning to get the best of both worlds and to address problems of scalability.
Cynthia Rudin gives a fantastic example of predicting imprisonment in 2 years. The US justice system (COMPAS) uses 137 features. She shows that the CORELS objective uses 4 rules and is just as accurate as COMPAS, trading 1% error for 1 less rule in the list.
Thanks to the organisers of the amazing
#dagstuhl
workshop on
#AIforsocialgood
. It was great (and fun!) sharing ideas on how AI can solve problems in global development for positive social impact.
This week in
#dagstuhl
: NGOs and computer scientists from 9 countries and 3 continents make AI advance the social good
orga: Claudia Clopath, Ruben De Winne
@rubendewinne
, Mohammad Emtiyaz Khan
@EmtiyazKhan
, Tom Schaul
PR:
Brendan Frey on AI to reprogram the human genome:
in ML4health gaining trust is important.
People complain AI is a blackbox. Interpretabiliity: we have to think of ways of explaining how things work.
We need generative models which also tell us what's not plausible
#NIPS2017
@kchonyc
@Khipu_AI
That was such an amazing tutorial - it was so clear and totally engaging. What an example. Thank you so much for your energy and for being a part of
#khipu2019
!
Just announced: the workshop on Climate informtics turns 10 next year and is moving from being a workshop to becoming a conference. Claire also gave a tutorial on this topic in 2014. You can check out the tutorial here:
#PeopleCentredcare
is such an important issue for
#HealthForAll
. It means shifting away our understanding of health and health systems from merely asking "What's the matter with you" and instead trying to understand from patients "What matters for you"
@catalinacangea
and Rhys will be presenting in person at the both, and I will be around virtually, so happy to grab a virtual coffee during these times to talk about
#ML4Healthcare
.
@s0f1ra
and
@weballergy
will also be around for virtual chats.
Beautifully articulated talk by
@CwaMaina
at
@black_in_ai
on the some of the kinds of problems we in the AI and ML community need to be thinking about at
#NIPS2017
During my past 2 months at , it has been amazing to be immersed in such an innovative space where we aim to push the boundaries of using AI to discover transformational medicines. I feel extremely fortunate to be part of such an amazing team.
@shamoons
@EmtiyazKhan
Great question - we prepared a blog last year on the decision process for inviting and selecting tutorials. We'll try to put some further guidance in the call for proposals. In the meantime, her's what
@aliceoh
and I wrote for the
#NeurIPS2019
tutorials
Really interesting perspective by
@cuttingforstone
at the
#ml4healthcare
conference talking about the need for empathy as researchers in the healthcare domain whose work also has a role in the function of caring for others.
This quote from Anatole Broyard strikes a chord. My father Bob Kale, who treated chronic pain patients, always told me that the his most effective treatment was spending time with his patients and listening to their stories
#mlhc2018
#ml4healthcare
Thanks
@cuttingforstone
!
This reminds me of when I first started working as an ML researcher in healthcare, I needed to do madatory training on the ethical principles for medical research and do a quiz on the declaration of Helsinki
At
#WiML2017
#NIPS2017
opening remarks being inspired by
@RaiaHadsell
sharing perspectives on her career path, on agents learning in a changing world, being successful lifelong learners and on embracing diversity. Wow!
@don_frank
The
#FOMO
is a lot worse when you're actually at
#NeurIPS2019
and you see tweets from people in all the other workshops and you want to be in all of the workshops at the same time (and attempt to do so!).
"Simple systems, simple models are the building blocks that help us to understand more complex systems" This is a talk filled with common sense that truly deserves the test of time award.
#NIPS2017