I have finally found time to start a blog. Although I don’t know if I will find the energy to do all this in the next few months, I am planning to use this blog in a variety of ways, with one post every first monday of the month:
- To develop in reasonable depth some ideas and concepts from machine learning and associated fields. This can go from general machine learning concepts (e.g., convex optimization) to more specialized topics (e.g., bias-variance trade-offs in stochastic gradient descent). The choice of topics will obviously be oriented towards my scientific interests and research.
- To present nice mathematical tools or tricks which I find cute, elegant and/or efficient, which I have encountered in recent years while reading (or writing) machine learning papers. The number of such tricks seems to be infinite and (spoiler alert for the students taking my classes) and are a good source of exam questions.
- To discuss any moderately research-oriented topics related to my various past and present roles as researcher, PhD advisor, reviewer, journal editor, conference area chair, conference organizer, etc.
Many other blogs which have been following for quite a while have been a source of inspiration and motivation: I am a bandit, hunch.net, Libre pensée.