#7 Paris Women in Machine Learning & Data Science : Recommender Systems, Optimization & Engineering Recruitment
For the last WiMLDS meet-up of the first season, we were 75 people to be hosted by Capital Fund Management in the beautiful Saint-Germain-des-Prés district of Paris.
The 7th meet-up of the Paris WiMLDS community started by a short presentation of Laurent Laloux, Chief Product Officer at Capital Fund Management. Laurent Laloux explained how CFM was involved in academia by financing several PhDs and supporting the Data Science Summer School at Polytechnique.
Before kicking-off the night, we started with exciting annoucements :
🌟Thanks to our meet-up, Jean-Marc found a PhD supervisor and Thai-An got an end-of-study internship in data science via our Slack channel!
🔜The Paris WiMLDS team will be attending the ICML conference, in Stockholm from the 10th to the 15th of July 2018, and will be represented by Chiara Biscaro, Data Scientist at Criteo Labs. We are happy to see that the ICML organization committee made sure women were represented in the program! This year, 50% of the keynote speakers are women.
🤝We are partnering with the Reactive Conference which will happen in Prague from the 29th to 31st of October 2018. You can get a 15% discount with the “ReactiveWiMLDS18" promo code!👈
We will start the second season of the Paris WiMLDS meet-ups on the 26th of September 2018 at BlaBlaCar with an amazing (yet secret) line-up!
Betty’s talk focused on the need for user-centered recommendation of relevant products, and on the challenges of finding an equilibrium between showing new products to discover or products similar to the one we already love and visited on the website. Different types of pages we visit also offer distinct opportunities for personalized recommendation, as what we see in a homepage is remarkably different from what’s instead in a product page. Both these setups present unique challenges and opportunities to show customers the best products for them to choose from.
If you want more, feel free to read the full article Betty Moreschini wrote.
Then, Pauline Tan, Post-Doc Researcher at CMLA (ENS Paris-Saclay) presented first order optimization methods applied to known image processing problems such as image denoising and deblurring. The methods proposed consisted of block-optimization where we divide the variable set into blocks, and then perform an alternate optimization on the direction of the block. She explained that such methods perform well in non-convex problems such as matrix factorization in machine learning. Finally Pauline presented a recent convergence result on non-convex energy functions where under a certain number of assumptions, they can show convergence to a local minimizer.
As always, this third part was interactive and invited the attendees to give feedback and share stories. If you missed the talk, Caroline Chavier wrote an article with even more details and references, or you can check her slides 👇
After the talks, we all enjoyed CFM’s beautiful terrace in central Paris. Everybody networked, discussed about Data Science & Machine Learning and admired the view!
The Paris WiMLDS team is already working on what’s next!
During the summer, you can :
🔗keep track of what’s going on thanks our Twitter account.
📩send an email to the Paris WiMLDS team to keep in touch >firstname.lastname@example.org
📍join our Slack channel for more discussions about machine learning, data science, and diversity in tech.
📑check our Google spreadsheet if you want to speak 📣, host 💙or help 🌠!
If you are not sure you want to attend the Paris WiMLDS meet-up, read one of our participants words about her experience! See you in September ☀️