Thank you and congratulations

Module 7. Evaluating model performance Wrap-up quiz
#1

Thank you and congratulations by this course
Thanks to everyone who made this course, it is a wonderful course with a great future too. I must say it was a pleasure to attend. The assistance of the teaching team is always very precise and clear.
As I am currently an engineer and interested in PhD, I wanted to ask you very kindly if you can guide me to apply for PhD and with future possibilities in the field.

#2

Thanks for the kind words!

As I am currently an engineer and interested in PhD, I wanted to ask you very kindly if you can guide me to apply for PhD and with future possibilities in the field.

So this is a broad question, and I am probably not the best person to answer (for example not being closely involved in research) … a generic recommendation would be to try to contact people closer to you (for example professors in your engineering school or alumni who have gone on to do a PhD in a domain you are interested in) to get a better feeling about what’s possible.

Another very generic recommendation is to follow your interests, maybe you can start applying what you have learned in this course in a Kaggle competition, maybe there are other more advanced or more domain-specific MOOCs that you can follow, maybe there is some open-source project you can get involved in? This kind of things may help making a difference in your CV.

In Inria, this happens regularly that you can have a temporary contract as an engineer say 6 months - 1 year and then start a PhD if things go well. This is a good opportunity to have some feeling what research actually looks like before doing a PhD.

#3

Another suggestion is to start reading the scientific publications around topics of your interest. Some familiarity with the literature is really helpful to apply for an MSc / PhD program and to have productive discussions with your future supervisor.

For machine learning, this is quite easy because all the most popular publication venues are Open Access and the rest is probably also available as pre-print on arxiv.org. Just browse the proceedings of the conference websites.

Some popular conferences for general machine learning:

  • ICML
  • NeurIPS
  • ICLR
  • AISTATS
  • COLT (much more on the theory side compared to the others)
  • JMLR (this one is a journal, the other are conferences)

For more applied domains:

  • CVPR
  • EMNLP
  • RECSYS

You can also find this website interesting: https://paperswithcode.com/

Also nowadays a lot of machine learning research and applications involve deep learning methods one way or another. So it would be a good idea to get familiar with deep learning and n

#4

Thank you so much, by the way regarding the kaggle how is this, how can I access it, is it online? Thanks

#5
#6

Thank you, in Kaggle also I could find some open-source project to see or maybe get involved in? Thanks

#7

Kaggle is a community of data scientists that share notebooks and compete on data science challenge in friendly way.

While kagglers typically use open source libraries, they do not necessarily maintain them. But still this is a good way .

I you are interested more specifically in contributing to open source you might be interested in attending sprints organized by organizations such as https://www.dataumbrella.org/, https://pydata.org/, https://numfocus.org/, https://conference.scipy.org/ and so on.