I wrote a Julia package (Shrike.jl) for approximate nearest neighbor search. My algorithm draws heavily on the mrpt algorithm and uses the same principles behind annoy (a popular approximate nearest neighbor package by Spotify). My Julia implementation be... (more…)

I've been more serious about learning Rust recently, after dragging on with passive learning for a while. My first real programming language was Julia, and I know other Julians interested in Rust. I've written this article for those people in mind, becaus... (more…)

Questions? Please register for JuliaCon: https://juliacon.org/2021/tickets/ and you will receive the link for Q/A via email. See you there!Julia has several ... (more…)

So, I participate in the Google Summer of Code 2020 with the project General partial differential equation solver using neural networks. (more…)

If you have a large collection of data and have to do similar computations on each element, data parallelism is an easy way to speedup computation using multiple CPUs and machines as well as GPU(s). While this is not the only kind of parallelism, it cover... (more…)