Put your money where your mouth is and build out this prototype.
This will be a YouTube video annotator. You watch a longer YouTube clip, you want to index it, get the meat out of it, so you can re-watch it yourself, or share a set of clips seamlessly to a friend.
This blog is about passing things along, as I’ve benefitted greatly from what others have offered on the internet. My degrees might as well have been from Google or John’s blog or Hacker News.
One-liner This paper studies and confirms the effectiveness of compositionality through human experiments from data generated by gaussian processes regression, supporting it’s case that human learning is inherently compositional. TLDR This paper uses eight carefully designed experiments to probe whether an inductive bias for compositionality exists in human learning Two class of theories around function learning exist: rule-based, parametric, strong inductive bias vs.
I don’t like the Yuketang interface. There’s many advantages to watching lectures playback. I like to watch video on VLC, where I can take notes and index parts of it. I got interested in this.
TLDR Lagrange multipliers is a constrained optimization method Lagrange multipliers restricts solutions to points that are feasible and stationary Key intuitive points: $\nabla{c}$ of our constraint function is normal to the surface.
The SFW (Safe for work) version What happens when you hand the steering wheel to the boy in the man? "Don’t think about what you want to be, but what you want to do.
Something I’ve been thinking a little about is how to learn more effectively. My main philosophy around this is DRY (Don’t Repeat Yourself) and KISS (Keep It Simple Stupid).
My friend David introduced me to LiquidText, an extremely useful iPad app for indexing and closely analyzing documents.
Because not all posts are publication-worthy. Not all ideas are fully baked, not all words are written, yet all trash should be put on the internet.
I just spent ten minutes trying to get an image asset on hugo to add the image I wanted.
Optimization Constrained Optimization Optimization is usually bounded. There is a taxonomy of that differ by:
type of objective linear non-linear quadratic type of constraints hard soft (optional with penalty) equality inequality number of variables and constraints Solving method slack variables (convert inequality to equality) penalty method with constraints optimal solution or approximation (global or local extrema) convexity differentiability Note that there are similarities and this grouping need not be mutually exclusive