mlr3: Pipelines

preview_player
Показать описание
  1. Statistical Learning and Data Science
Рекомендации по теме
mlr3: Pipelines 0:37:58

mlr3: Pipelines

mlr3: Introduction 0:20:23

mlr3: Introduction

Slangbos Mapping - 0:07:28

Slangbos Mapping - Introduction to mlr3 for machine-learning algorithms in R

mlr3: Tuning 0:46:08

mlr3: Tuning

Why R? Webinar 1:28:06

Why R? Webinar 009 - Bernd Bischl, Florian Pfisterer, Martin Binder - Pipelines and AutoML with mlr3

mlr3: Resampling 0:26:27

mlr3: Resampling

I2ML - Tuning 0:20:02

I2ML - Tuning - Pipelines and AutoML

R : mlr3 0:01:39

R : mlr3 PipeOps: Create branches with different data transformations and benchmark different learne

useR! 2020: Machine 3:03:01

useR! 2020: Machine Learning with mlr3 (Bernd Bischl, Michel Lang), tutorial

Introduction to mlr3 1:10:34

Introduction to mlr3 for Machine Learning with R Webinar

useR! 2019 Toulouse 0:15:28

useR! 2019 Toulouse - Talk Data Mining - Bernd Bischl

**2022-10-26 (meeting #103)** 0:53:53

**2022-10-26 (meeting #103)** Intro to the {mlr3} package in R (Ed)

Session 3A - 1:00:43

Session 3A - Machine Learning and Data Management

R : Where 0:00:56

R : Where does mlr3 save the final model?

Session 5B - 1:04:59

Session 5B - Mathematical Models

I2ML - Nested 0:07:06

I2ML - Nested Resampling

Towards an Enterprise 1:22:22

Towards an Enterprise Grade Machine Learning Pipeline with R

SPb R User 1:32:19

SPb R User group meetup. International

useR! 2019 Toulouse 0:18:35

useR! 2019 Toulouse - Talk Data Mining - Michel Lang

Optimization in Machine 0:27:34

Optimization in Machine Learning - Evolutionary Algorithms - Introduction

R - простой 0:37:59

R - простой пример задачи классификации с mlr3. Kaggle titanic dataset...

Manning Introduces - 0:01:57

Manning Introduces - Machine Learning with R, tidyverse, and mlr

OpenML + R 0:41:17

OpenML + R + mlr

Комментарии
Автор

Seeing how you can build any thinkable pipe, I was thinking that building a automl-system like "TPOT" would actually now be possible in the R world. Probably even something better!
The idea is to optimizie the pipeline with a genetic algorithm, like the ones the GA package provides. So I guess the GA would choose between different pipelines and then also inside of these pipelines between different hyperparameters.

BlueQualityRhythm