I’ve been interested in participating in data science and machine learning competitions for some time. And I believe Julia can be a great language in these competitions.
I understand that Kaggle seems to be the most popular platform. Although it still doesn’t support Julia in its platforms, there are competitions where you can use the language (featured competitions). Additionally, there are other platform alternatives like DrivenData, where one is free to participate in the language of their choice. However, this comes at the cost of not having a large community or many opportunities to share code.
That said, I wanted to ask if you’ve had experiences using Julia for these competitions, on which platforms and competitions you’ve participated, the positive and negative experiences of using the language, or whether you believe it has provided you with any advantage over other competitors using Python or R. Learning about instances where Julia has genuinely excelled would be truly motivating.
I usually prepare and participante in competions. Because my courses we use Python, I use it, bit also I use Julia also for testing. I love DataFrame more than Pandas. However, for Kaggle competitions, you only need submit the solutions so you can use Julia (not for notebooks published in Kaggle). In muy opinion, Julia is almost better for exploiring the data. MLJ is rather good, very competitive in its nice API. Try it!
Great to know that you’re using it for competitions!
But not all competitions make that possible. If you go to Kaggle today, all competitions tagged with “code competition” require you to upload the associated notebook when making submissions. Therefore, you have very few options where you can really use Julia.
In fact, I saw a couple of notebooks on the platform with Julia code, like this example that creates an animation with Makie and another one with Flux. This is just markdown, but I think it’s quite effective for showcasing use cases.
I really believe that Julia is becoming increasingly competitive, but I was still interested in knowing if there are people actively competing and using Julia on these platforms today.