Building Machine Learning Pipelines in PySpark MLlib

4.5
14件の評価
提供:
Coursera Project Network
このガイド付きプロジェクトでは、次のことを行います。

Learn how to create a Random Forest pipeline in PySpark

Learn how to choose best model parameters using Cross Validation and Hyperparameter tuning in PySpark

Learn how to create predictions and assess model's performance in PySpark

Clock1.5 hours
Intermediate中級
Cloudダウンロード不要
Video分割画面ビデオ
Comment Dots英語
Laptopデスクトップのみ

By the end of this project, you will learn how to create machine learning pipelines using Python and Spark, free, open-source programs that you can download. You will learn how to load your dataset in Spark and learn how to perform basic cleaning techniques such as removing columns with high missing values and removing rows with missing values. You will then create a machine learning pipeline with a random forest regression model. You will use cross validation and parameter tuning to select the best model from the pipeline. Lastly, you will evaluate your model’s performance using various metrics. A pipeline in Spark combines multiple execution steps in the order of their execution. So rather than executing the steps individually, one can put them in a pipeline to streamline the machine learning process. You can save this pipeline, share it with your colleagues, and load it back again effortlessly. Note: You should have a Gmail account which you will use to sign into Google Colab. Note: This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.

あなたが開発するスキル

Machine Learning Pipelineshyperparameter tuningPySparkCross Validation

ステップバイステップで学習します

ワークエリアを使用した分割画面で再生するビデオでは、講師がこれらの手順を説明します。

  1. Install Spark on Google Colab and load a dataset in PySpark

  2. Describe and clean your dataset

  3. Create a Random Forest pipeline to predict car prices

  4. Create a cross validator for hyperparameter tuning

  5. Train your model and predict test set car prices

  6. Evaluate your model’s performance via several metrics

ガイド付きプロジェクトの仕組み

ワークスペースは、ブラウザに完全にロードされたクラウドデスクトップですので、ダウンロードは不要です

分割画面のビデオで、講師が手順ごとにガイドします

レビュー

BUILDING MACHINE LEARNING PIPELINES IN PYSPARK MLLIB からの人気レビュー

すべてのレビューを見る

よくある質問

よくある質問

さらに質問がある場合は、受講者向けヘルプセンターにアクセスしてください。