Amazon Interview

 Posted on April 14, 2020  |  3 minutes  |  524 words  |  John Chen

Addicted to interviewing. Just finished the Amazon interview. It was a shadowing interview, so there was a bonus interviewer! Surprisingly, it was not a coding challenge! Instead they asked (quite a few) questions about my machine learning knowledge. Things I did well: Answered the first LP question nicely (overall). Tell me about a time when you went outside your scope to improve something. Should have mentioned I got an award for that. [Read More]

Vector Compute Guide

 Posted on March 22, 2020  |  1 minutes  |  44 words  |  John Chen

Vector has two computing servers. One is the MaRS cluster, while the other is the Vaughan cluster.

You can access the Vaughan cluster, via vremote.vectorinstitute.ai You can access the MaRS cluster, vi q.vectorinstitute.ai

The power of e-mail inboxing. And making sure things get done!

Elements of Web Programming

 Posted on March 14, 2020  |  1 minutes  |  176 words  |  John Chen

I have long believed that Web programming is one of the most difficult parts of practical, software development. As the name suggests, full-stack web development encapsulates all levels of programming, from database to backend to frontend. Not only is there breadth across the stack, but there is also breadth across the different types of stacks. For instance, PHP web development is completely different from modern Node ES6 development is completely different from C# web development. [Read More]

Introduction to Group Theory

 Posted on March 7, 2020  |  2 minutes  |  250 words  |  John Chen

I am becoming more interested in rigorizing my knowledge of mathematics. A group is a set of elements $$G$$ and an operation $$\bigotimes$$. It satisfies the following four properties: Closure Associativity Neutral Element Inverse Element From Groups to Vector Spaces: If we enhance our group with an outer operation, then we get a vector space! Roughly, this outer operation takes an element from inside our set, and an external element from a field (such as the reals) and then gives us a way of combining the inner element with the external element. [Read More]

We are all going south

 Posted on February 19, 2020  |  2 minutes  |  335 words  |  John Chen

Attended the Rich Sutton talk today at Vector. It was a great talk, where he soliloquized about his research agenda and motivation, and used this to motivate his work on SuperDyna, a general intelligence framework based on his work on options. At the end of this post are some rough notes from the presentation. But I wanted to discuss the key takeaways I got. I had the opportunity to talk with him personally briefly for a few moments as well over lunch. [Read More]

Making Imperfect Decisions

 Posted on January 27, 2020  |  2 minutes  |  300 words  |  John Chen

Sorry for the long hiatus everyone. I am refraining from blogging so much while I figure out some infrastructure related details: i.e. how to write inline latex effectively in the blog, how to use pictures, and importantly how to categorize things by date, etc. so that this doesn’t blow up. (In partciular, you can already see how the number of posts will grow linearly with time: not good). The paradox of graduate school: learning how to make imperfect decisions So now I have been in grad school for over a semester. [Read More]

Regressions

 Posted on January 20, 2020  |  1 minutes  |  90 words  |  John Chen

There is a powerful, and statistical way of thinking about all regressions. Instead of viewing regression as simply “minimizing the least squares error”, or some type of MLE, we can instead think about the statistical, generative model of our process. This is the approach to regression that talks about $\Beta$, normal distribution, and the residuals are normally distributed! Essentially, Poisson regression simply makes a different modelling assumption/perspective: that we look at a $ln$ viewpoint! [Read More]

Adobe

 Posted on January 8, 2020  |  1 minutes  |  191 words  |  John Chen

Remember when I said interviewing season was over? Well, not quite. Just finished the interview with Adobe. This was one of the more frazzled experiences, and it did not go as planned. To start off, I forgot my laptop at home, and this caused me to almost be late for the interview. Second, I kept on having connection problems during the interview. I was constantly disconnecting from the collabedit, which made it hard to concentrate/focus. [Read More]

Research Log

 Posted on January 7, 2020  |  1 minutes  |  94 words  |  John Chen

Worked on the RL project again today. I implemented: 1. More updates (to the policy network) 2. Dynamic reward baselining, with rolling average of past rewards I am also working in the oracle clusters paradigm. If we know the actual labels, we should be able to pick optimally. Yet somehow, we are unable to. We keep falling into policy collapse, due to the attractors (local minima) in the RL process. [Read More]