First Post - What does this blog stand for?

Hello,

as you can see, this is the first post of this blog. I'll use to introduce the blog, myself, and so on...

I'm a first year PhD student at the Department of Computing Science at the University of Alberta. Previously, I took my B.Sc. and M.Sc. degrees in Brazil, at the Universidade Federal de Minas Gerais (UFMG), also in Computer Science.  Yes, I'm Brazilian and I do love soccer.  I root for Cruzeiro but this is not really relevant for this discussion.

This blog exists because I've read a lot about how to succeed in grad school before starting the PhD, and a common advise was to write, write and write... Since I'm not an English native-speaker, I suppose I have to write even more, so I'll use this blog for this. I cannot guarantee that every text will be perfect, and if you notice a typo, a not-so-good-written text or an error, please let me know.

What am I going to write about? Most of the part about Artificial Intelligence, Machine Learning and research. I struggle to create a classification between AI and ML, sometimes it seems redundant to mention both, but whatever... This is what this blog is about.

While reading texts about being a successful PhD student, I always thought that one of my first posts would be related to this subject, summarizing everything that I read. However, a friend, also a former student at UFMG and now a PhD Student at UCSB, Arlei Silva, did it very well [link]. So it is useless to write about the same thing again.

Thus, my first posts are probably going to discuss the main concepts of the online learning field (not e-learning!). To the reader not used to this concept, I will finish with a quote extracted from the website of the online learning course at the University of Alberta, taught by professor Csaba Szepesvari:

"Online learning is a hot topic in machine learning: When processing huge amounts of data algorithms must operate continuously in their environments, making predictions, receiving rewards, suffering losses. The distinguishing feature of online problems is that the algorithm's performance is measured by comparing it with the best performing algorithm from a larger class of algorithms, allowing for a wide class of environment models, ranging from stationary stochastic models to ones where no statistical constraints are adopted."

I hope you enjoy the blog. First, I hope some people read it, to avoid it becoming some kind of my public diary. Disclaimer: I'm a PhD student, I may not be able to update this blog with the frequency I would like to do so.