[{"data":1,"prerenderedAt":182},["ShallowReactive",2],{"topic-page:\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Fsupervised-unsupervised-and-reinforcement-learning":3},{"topic":4,"category":142,"seo":146,"breadcrumbs":151,"prerequisites":161,"nextTopics":162,"relatedTopics":170},{"id":5,"slug":6,"title":7,"summary":8,"canonicalPath":9,"categoryPath":10,"difficulty":11,"targetAudience":12,"estimatedMinutes":13,"generatedAt":14,"publishedAt":14,"updatedAt":14,"reviewStatus":15,"tags":16,"expectedKnowledge":25,"learningGoal":28,"keyTerms":29,"blocks":55},"it-aws-aws-certified-ai-practitioner-ai-ml-fundamentals-supervised-unsupervised-and-reinforcement-learning","supervised-unsupervised-and-reinforcement-learning","教師あり・教師なし・強化学習","教師あり・教師なし・強化学習を、AIF-C01の公式objective 1.1.5に沿って、用語、判断順序、例、誤解修正で整理します。","\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Fsupervised-unsupervised-and-reinforcement-learning","\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals","beginner","cert_candidate",6,"2026-08-06","ai_generated",[17,18,19,20,21,22,23,24],"AWS Certified AI Practitioner","AIF-C01","AI・機械学習の基礎","教師あり学習","教師なし学習","強化学習","ラベル","報酬",[26,27],"AWSクラウドの基本用語を聞いたことがある","AIを使う業務例を日常的な言葉で考えられる","教師あり・教師なし・強化学習について、試験で問われる違いと選択基準を説明し、短い業務例へ適用できる。",[30,34,37,40,43,46,50],{"id":31,"label":20,"shortDefinition":32,"role":33},"supervised-learning","入力と正解ラベルの組から、未知入力への予測規則を学ぶ方式です。","primary",{"id":35,"label":21,"shortDefinition":36,"role":33},"unsupervised-learning","正解ラベルを使わず、データのまとまりや構造を見つける方式です。",{"id":38,"label":22,"shortDefinition":39,"role":33},"reinforcement-learning","環境で行動し、将来を含む報酬が大きくなる方策を学ぶ方式です。",{"id":41,"label":23,"shortDefinition":42,"role":33},"label","教師あり学習で、入力に対応する正解や分類名として与える値です。",{"id":44,"label":24,"shortDefinition":45,"role":33},"reward","強化学習で、行動結果の望ましさを数値で示す学習信号です。",{"id":47,"label":48,"shortDefinition":49,"role":33},"policy","方策","強化学習で、状態に応じてどの行動を選ぶかを定める規則です。",{"id":51,"label":52,"shortDefinition":53,"role":54},"clustering","クラスタリング","正解ラベルなしで、似た特徴を持つデータを群へまとめる手法です。","contextual",[56,61,72,81,90,96,102,109,123,132],{"id":57,"type":58,"title":7,"subtitle":59,"shortDefinition":60},"hero","HeroBlock","正解ラベル、隠れたまとまり、報酬のどれを使うか","モデルが経験から規則性を得るとき、与える手掛かりを正解ラベル、データの構造、行動への報酬に分けた考え方です。",{"id":62,"type":63,"term":20,"definition":64,"plainExplanation":65,"presentation":66,"supportLabel":67,"keywords":68},"definition","DefinitionBlock","[[cloze:supervised-learning|教師あり学習]]とは、入力と正解ラベルの組から、未知入力への予測規則を学ぶ方式です。","この方式は入力と正解の組を使い、未知入力の値やクラスを予測します。[[term:unsupervised-learning|教師なし学習]]は正解[[term:label|ラベル]]を与えず、似た顧客のまとまりや低次元の構造を見つけます。[[term:reinforcement-learning|強化学習]]は環境で行動し、将来を含む[[term:reward|報酬]]が大きくなる[[term:policy|方策]]を学びます。名称の『教師』は人が常時指示する意味ではなく、学習データに正解があるかを表します。","natural","見分けるポイント",[69,70,71,24],"正解ラベル","構造の探索","行動",{"id":73,"type":74,"steps":75,"title":80},"decision-steps","StepBlock",[76,77,78,79],"予測したい正解が過去データに付いているかを見る。","正解なしで自然な群や構造を探す目的か確認する。","連続する行動と、その結果への報酬があるかを見る。","目的が予測、探索、方策のどれかを言葉にして選ぶ。","学習方式を見抜く順序",{"id":82,"type":83,"title":84,"points":85,"body":89},"exam-cues","KeyPointBlock","判断で押さえる境界",[86,87,88],"正解ラベルありは教師あり","ラベルなしの構造探索は教師なし","行動と累積報酬は強化学習","同じ業務でも目的で方式が変わります。既知の解約者ラベルから次の解約を当てるのは教師あり、購買傾向だけから顧客群を発見するのは教師なし、在庫補充の行動を長期利益で改善するのは強化学習です。アルゴリズム名を先に選ばず、学習時に得られる信号を見ます。",{"id":91,"type":92,"src":93,"alt":94,"caption":95},"diagram","DiagramBlock","\u002Fassets\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Fsupervised-unsupervised-and-reinforcement-learning\u002Fsupervised-unsupervised-and-reinforcement-learning.svg","教師あり・教師なし・強化学習について教師あり：正解、教師なし：構造、強化学習：報酬、出力：予測・群・方策の関係を文字ラベルと矢印で示した図","学習時に何を手掛かりとして受け取るかを並べた比較図です。業務名ではなく、正解・構造・報酬で見分けます。",{"id":97,"type":98,"scenario":99,"explanation":100,"result":101},"worked-example","ExampleBlock","通販会社が顧客分析を三つの目的で行います。","過去の不正ラベルから新規取引を判定するなら教師あり、ラベルなしで顧客を購買パターン別に分けるなら教師なし、割引提示を繰り返しながら長期利益を高める行動を学ぶなら強化学習です。データは同じ顧客情報でも、学習信号が違います。","問題文の『正解』『似た群』『行動後の報酬』を目印に、方式を一意に絞れます。",{"id":103,"type":104,"warningTitle":105,"message":106,"severity":107,"action":108},"misconception","WarningBlock","クラスタリング結果は正解ラベルではない","[[term:clustering|クラスタリング]]が作る群には、初めから『優良顧客』などの正解名が付いているとは限りません。分析後に人が特徴を読み、業務上の意味を付けます。","warning","データに何が付いているかではなく、学習に正解として渡す情報は何かを確認します。",{"id":110,"type":111,"title":112,"questions":113},"quiz","QuizBlock","AIF-C01 判断チェック",[114],{"question":115,"choices":116,"answerIndex":118,"choiceExplanations":119},"正解ラベルのない利用履歴から似た行動の顧客群を探す方式はどれですか。",[21,20,117],"報酬に基づく強化学習",0,[120,121,122],"正解です。ラベルなしの構造を探します。","予測対象の正解ラベルがありません。","行動と報酬の反復が示されていません。",{"id":124,"type":125,"title":126,"points":127},"summary","SummaryBlock","まとめ",[128,129,130,131],"教師ありは入力と正解の組を使う","教師なしはラベルなしの構造を探す","強化学習は行動と報酬から方策を学ぶ","業務名より学習信号を見る",{"id":133,"type":134,"title":135,"nextTopics":136,"relatedKeywords":138},"next-action","NextActionBlock","次は「AIの価値と使わない判断」へ",[137],"it-aws-aws-certified-ai-practitioner-ai-ml-fundamentals-ai-business-value-and-non-ai-boundaries",[139,140,141],"AI適用判断","費用対効果","自動化",{"path":10,"title":19,"description":143,"parentPath":144,"accentToken":145},"AI、機械学習、学習方式、代表ユースケース、開発ライフサイクルを、AIF-C01の判断問題に沿って整理します。","\u002Fit\u002Faws\u002Faws-certified-ai-practitioner","cloud-amber",{"title":147,"description":148,"canonicalUrl":149,"ogImage":150},"教師あり・教師なし・強化学習 | AWS Certified AI Practitioner","AIF-C01対策として教師あり・教師なし・強化学習の定義、判断軸、具体例、誤解しやすい境界を公式試験範囲に沿って学びます。","https:\u002F\u002Fzeqnilo.com\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Fsupervised-unsupervised-and-reinforcement-learning","https:\u002F\u002Fzeqnilo.com\u002Fassets\u002Fsite\u002Fdefault-og.svg",[152,155,158,159,160],{"title":153,"path":154},"IT","\u002Fit",{"title":156,"path":157},"AWS","\u002Fit\u002Faws",{"title":17,"path":144},{"title":19,"path":10},{"title":7,"path":9},[],[163],{"id":137,"title":164,"summary":165,"canonicalPath":166,"categoryPath":10,"categoryTitle":19,"difficulty":11,"targetAudience":12,"estimatedMinutes":13,"publishedAt":14,"updatedAt":14,"reviewStatus":15,"tags":167},"AIの価値と使わない判断","AIの価値と使わない判断を、AIF-C01の公式objective 1.2.1・1.2.2に沿って、用語、判断順序、例、誤解修正で整理します。","\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Fai-business-value-and-non-ai-boundaries",[17,18,19,139,140,141,168,169],"不確実性","事業価値",[171],{"id":172,"title":173,"summary":174,"canonicalPath":175,"categoryPath":10,"categoryTitle":19,"difficulty":11,"targetAudience":12,"estimatedMinutes":13,"publishedAt":14,"updatedAt":14,"reviewStatus":15,"tags":176},"it-aws-aws-certified-ai-practitioner-ai-ml-fundamentals-inference-modes-and-data-types","推論方式とデータの種類","推論方式とデータの種類を、AIF-C01の公式objective 1.1.3・1.1.4に沿って、用語、判断順序、例、誤解修正で整理します。","\u002Fit\u002Faws\u002Faws-certified-ai-practitioner\u002Fai-ml-fundamentals\u002Finference-modes-and-data-types",[17,18,19,177,178,179,180,181],"推論","バッチ","リアルタイム","非同期","データ型",1787370539570]