Merge pull request #46 from Evovest/dev

Dev
Evovest · Mar 13, 2020 · 4cd2e9e · 4cd2e9e
2 parents 44c37ec + 598a4e8
commit 4cd2e9e
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 3 deletions.
diff --git a/Project.toml b/Project.toml
@@ -15,7 +15,7 @@ StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 [compat]
 CategoricalArrays = "0.7"
 Distributions = "0.22"
-MLJModelInterface = "0.1"
+MLJModelInterface = "0.1, 0.2"
 StaticArrays = "0.12"
 StatsBase = "0.32"
 julia = "1"

diff --git a/README.md b/README.md
@@ -3,11 +3,10 @@
 [![Build Status](https://travis-ci.org/Evovest/EvoTrees.jl.svg?branch=master)](https://travis-ci.org/Evovest/EvoTrees.jl)
 
 A Julia implementation of boosted trees.
+Efficient histogram based algorithm with support for conventional and extended loss (notably multi-target objectives such as max likelihood methods).
 
 [R binding available](https://github.com/Evovest/EvoTrees)
 
-Provides flexibility for efficient custom objectives (notably multi-target objectives such as max likelihood methods).
-
 Currently supports:
 
 - linear
@@ -18,6 +17,7 @@ Currently supports:
 - multiclassification (softmax)
 - Gaussian (max likelihood)
 
+
 Input features is expected to be `Matrix{Float64}`. User friendly format conversion to be done.
 Next priority: GPU support.
 
@@ -35,6 +35,18 @@ Official Repo:
 julia> Pkg.add("EvoTrees")
 ```
 
+
+## Performance
+
+[Benchmark](https://github.com/Evovest/EvoTrees.jl/blob/master/blog/benchmarks.jl) for 100 iterations on randomly generated data:
+
+| Dimensions / Algo | XGBoost Exact | XGBoost Hist | EvoTrees |   |
+|-------------------|:-------------:|:------------:|:--------:|---|
+| 10K x 100         |     1.18s     |     2.15s    |   0.52s  |   |
+| 100K x 100        |     9.39s     |     4.25s    |   2.02s  |   |
+| 1M X 100          |     146.5s    |     20.2s    |   22.5   |   |
+
+
 ## Parameters
 
   - loss: {:linear, :logistic, :poisson, :L1, :quantile, :softmax, :gaussian}