Ayush Patnaik

Brilliant!

https://juliastats.org/Survival.jl/latest/km/#StatsAPI.confint-Tuple{KaplanMeier}

@mousum-github

What's left in the PR @ArunSanganal?

@codetalker7 let's do this once you are available. It'll help us understand MLJ.

I attempted Bayesian optimisation for this, and at least for some problems, it turned out to be a bit better. ```julia function smoothcv(λ) n = length(X) k = n avgrss...

> In `smoothcv`, the arguement `λ` - is it scalar? or is it vector? From your code, I think it is scalar. However, in the following line `λ` is trated...

I took a sample of 1000 rows. I didn't run `@btime` but `@time`, so just a single run instead of the average of a few runs: ```julia julia> @time bayes_optimization(smoothcv,...

Hi Sourish, I found the book. It's a really neat trick. When you build the model, you can get LOOCV score using the intermediate calculations. This is the reason LOOCV...

How about StatsModels API also? So, we'll have ```julia fit(Loess, @formula(y ~ x), dataframe; span = 0.5) ``` Or / and ```julia Loess(@formula(y ~ x), dataframe; span = 0.5) ```