fd-update

__site/assets/literate/A-fit-predict.md

@@ -7,7 +7,7 @@
 
 ### Data
 
-As in "[choosing a model](choosing-a-model.html)", let's load the Iris dataset and unpack it:
+As in "[choosing a model](/getting-started/choosing-a-model/)", let's load the Iris dataset and unpack it:
 
 ```julia:ex1
 using MLJ, Statistics, PrettyPrinting

__site/assets/literate/D0-loading.md

@@ -26,7 +26,7 @@ The `dataset` function returns a `DataFrame` object from the [DataFrames.jl](htt
 typeof(boston)
 ```
 
-For a short introduction to DataFrame objects, see [this tutorial](/pub/data/dataframe.html).
+For a short introduction to DataFrame objects, see [this tutorial](/data/).
 
 ## Using CSV
 

__site/assets/literate/ISL-lab-3.md

@@ -32,7 +32,7 @@ describe(boston, :mean, :std, :eltype)
 ```
 
 So there's no missing value and most variables are encoded as floating point numbers.
-In MLJ it's important to specify the interpretation of the features (should it be considered as a Continuous feature, as a Count, ...?), see also [this tutorial section](/pub/getting-started/choosing-a-model.html#data_and_its_interpretation) on scientific types.
+In MLJ it's important to specify the interpretation of the features (should it be considered as a Continuous feature, as a Count, ...?), see also [this tutorial section](/getting-started/choosing-a-model/#data_and_its_interpretation) on scientific types.
 
 Here we will just interpret the integer features as continuous as we will just use a basic linear regression:
 

__site/assets/literate/ISL-lab-5.md

@@ -136,7 +136,7 @@ savefig(joinpath(@OUTPUT, "ISL-lab-5-g3.svg")) # hide
 
 Let's crossvalidate over the degree of the  polynomial.
 
-**Note**: there's a  bit of gymnastics here because MLJ doesn't directly support a polynomial regression; see our tutorial on [tuning models](/pub/getting-started/model-tuning.html) for a gentler introduction to model tuning.
+**Note**: there's a  bit of gymnastics here because MLJ doesn't directly support a polynomial regression; see our tutorial on [tuning models](/getting-started/model-tuning/) for a gentler introduction to model tuning.
 The gist of the following code is to create a dataframe where each column is a power of the `Horsepower` feature from 1 to 10 and we build a series of regression models using incrementally more of those features (higher degree):
 
 ```julia:ex11