Pipelines and Feature Unions in Scikit Learn

Показать описание

We talk about the most power features of scikit learn: pipelines and feature unions (combining estimators to create even more powerful ones)

Associated Github Commit:

Associated Scikit Links:

Рекомендации по теме

Комментарии

One quick suggestion can you please increase the font

ravimahar

Data Talks
Pipelines and Feature Unions (sci-kit learn)
00:00
We're talking about a little bit of data science and scikit-learn where this time we learn just a tiny amount of data science and a whole lot of scikit-learn. Let's get started. This is my favorite part of scikit-learn. This is one of the coolest parts about it. 0:16

fet

1:04
I named them <reduce_dim> (dimensionality_reduction) and <cls> (classification). I feed them into a pipeline, and tada. I get the entire thing. So it's collapsing both of them into one. It fits the PCA, then applies the transformation from the PCA. And then it fits the classifier and then applies the prediction from the classifier. (1:21)

fet

how do identify which features came from which transformation ?

rashmimahadevaiah

0:43
The only parameter it takes is a LIST of ESTIMATORS, a list of steps. Let me just show you what this looks like. So, for example, I've got two estimators here. I've got a PCA and I've got an SVC. And so of a dimensionality reduction and then classification. I named them, you've to name them - this is part of it. (1:04)

fet

Hi
Thank you for this video.
Does anyone has an example where feature union outperforms the single feature selection approach?

zapy

0:16
It's called pipelines and feature unions. So let's just get off and you can sort of see why I think it's so cool. So what is a pipeline? A pipeline is a list of transformations. We've seen estimators, we've seen meta-estimators, which take estimators to make new estimators. Pipelines take a list of estimators to make a whole new estimator. They just squash it all into 1 estimator. 0:43

fet

Why does your pipe have a classifier, and your params have 2 classifiers?

davidaloha

Pipelines and Feature Unions in Scikit Learn

Pipelines and Feature Unions in Scikit Learn

Composite Estimators using Pipelines & Feature Unions - Machine learning | Learnbay

Transformers, Estimators, and Pipelines (3/5)

SKLEARN PIPELINE AVANCÉE

Understanding Pipeline in Machine Learning with Scikit-learn (sklearn pipeline)

Introduction to Scikit-Learn pipeline API

Sklearn pipelines | Emily Godson

Composite Estimators using Pipelines & FeatureUnions

Database Objects | Databricks Master Series 2025 | Video Podcast | Prajesh Jha

Bugra Akyildiz - A Thorough Machine Learning Pipeline via Scikit Learn

A Framework for Feature Engineering and Machine Learning Pipelines - Jacques Peeters

Machine Learning Pipeline - Step 3 - Feature Selection

SK Learn Pipelines

Four ways to examine the steps of a Pipeline

Access part of a Pipeline using slicing

From Pipelines to Refineries: Building Complex Data Applications with Apache Spark - Tim Hunter

Learn how practically use Pipeline & column transformers in Machine learning (Sklearn)

Pipeline in Machine Learning | sklearn | TeKnowledGeek

Pandas Pipes and Scikit-Learn Pipelines Tutorials (Hepatitis Prediction)

Creating Pipelines Using SKlearn- Machine Learning Tutorial

ITS 365 - The Machine Learning Pipeline and Feature Engineering

Type Safe Data Processing and Machine Learning Pipelines with Flyte and Pandera | Union

Scikit Learn Column Transformers - Pipeline

Designing data pipelines for analytics and machine learning in industrial settings