{"templateId":"markdown","sharedDataIds":{"sidebar":"sidebar-@l10n/ja/sidebars.yaml"},"props":{"metadata":{"markdoc":{"tagList":[]},"type":"markdown"},"seo":{"title":"機械学習パイプラインの管理","description":"Treasure Data Product Documentation · Collect and Unify · Segment and Activate · Experiment and Analyze · Decisioning Automate with AI Scale and Trust.","siteUrl":"https://docs.treasuredata.com","lang":"en-US","llmstxt":{"hide":false,"sections":[{"title":"Table of contents","includeFiles":["**/*"],"excludeFiles":[]}],"excludeFiles":[]}},"dynamicMarkdocComponents":[],"compilationErrors":[],"ast":{"$$mdtype":"Tag","name":"article","attributes":{},"children":[{"$$mdtype":"Tag","name":"Heading","attributes":{"level":1,"id":"機械学習パイプラインの管理","__idx":0},"children":["機械学習パイプラインの管理"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Treasure Dataにデータを取り込んだ後、Treasure Dataのクエリとワークフローを使用して予測モデルを構築できます。"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["教師あり学習の典型的な機械学習パイプラインは次のとおりです:"]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["データ準備"]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["モデルの構築"]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["モデルの評価"]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["学習済みモデルによる未知のデータの予測"]}]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Treasure ワークフローを使用して、教師あり学習プロセスを管理できます。Treasure Data (TD) は、使い慣れた",{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/products/customer-data-platform/data-workbench/workflows"},"children":["Treasure ワークフロー"]},"環境内で設定できる機能としてAutoMLを提供しています。",{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/products/customer-data-platform/machine-learning/automl"},"children":["AutoML"]},"の詳細をご確認ください。"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"img","attributes":{"src":"/assets/105578848.2775c3c343559d67e1778594d2fc1c2403686c7398feafe43ccbdbe8a0ad1bce.3cb60505.png","alt":""},"children":[]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Treasure ワークフロー内でDigdag Treasure Dataオペレータを使用することで、データ準備から予測まで機械学習を自動化できます。Digdag Treasure Dataオペレータには次のものが含まれます:"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td.html"},"children":["td>: Treasure Dataクエリ"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_run.html"},"children":["td_run>: Treasure Data保存済みクエリ"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_ddl.html"},"children":["td_ddl>: Treasure Data操作"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_load.html"},"children":["td_load>: Treasure Dataバルクローディング"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_for_each.html"},"children":["td_for_each>: Treasure Dataクエリを使用した繰り返し"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_wait.html"},"children":["td_wait>: Treasure Dataテーブルへのデータ到着の待機"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_wait_table.html"},"children":["td_wait_table>: Treasure Dataテーブルへのデータ到着の待機"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_partial_delete.html"},"children":["td_partial_delete>: Treasure Dataテーブルの範囲削除"]}]}]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"https://docs.digdag.io/operators/td_table_export.html"},"children":["td_table_export>: Treasure DataテーブルのS3へのエクスポート"]}]}]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Digdagはタスクを並列に実行できるため、パラメータチューニングなどの独立したタスクを同時に実行できます。Treasure ワークフローを使用すると、予測タスクを製品提供の定期的な部分にすることができます。時間単位または日単位でバッチで機械学習プロセスを実行および進化させる安定した方法を持つことは、それらを進化させ、より優れた予測モデルを導き出すための良い方法です。"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"img","attributes":{"src":"/assets/105578847.ce353f50a12e39ebe398e8039749d93187b5c9e549cbd0c65abfa050a1499f71.3cb60505.png","alt":""},"children":[]}]}]},"headings":[{"value":"機械学習パイプラインの管理","id":"機械学習パイプラインの管理","depth":1}],"frontmatter":{"seo":{"title":"機械学習パイプラインの管理"}},"lastModified":"2026-06-01T09:47:58.000Z","pagePropGetterError":{"message":"","name":""}},"slug":"/ja/products/customer-data-platform/machine-learning/automl/managing-the-machine-learning-pipeline","userData":{"isAuthenticated":false,"teams":["anonymous"]},"isPublic":true}