이 단원에서는 숫자처럼 동작하는 정수 또는 부동 소수점 값을 의미하는 숫자 데이터에 중점을 둡니다. 즉, 덧셈, 집계, 순서 지정 등이 가능합니다. 다음 단원에서는 카테고리처럼 작동하는 숫자가 포함될 수 있는 범주형 데이터에 중점을 둡니다. 세 번째 단원에서는 모델을 학습하고 평가할 때 고품질 결과를 얻기 위해 데이터를 준비하는 방법을 중점적으로 다룹니다.
수치 데이터의 예는 다음과 같습니다.
온도
무게
자연 보호 구역에서 겨울을 나는 사슴의 수
반대로 미국 우편번호가 5자리 또는 9자리 숫자인 경우, 숫자처럼 동작하거나 관계라고 할 수 있습니다. 우편번호 40004 (켄터키주 넬슨 카운티)는 워싱턴 D.C.의 우편번호 20002의 두 배가 아닙니다. 이러한 숫자는 카테고리, 특히 지리적 지역을 나타내며 범주형 데이터로 간주됩니다.
[null,null,["최종 업데이트: 2025-07-27(UTC)"],[[["\u003cp\u003eThis module focuses on preparing numerical data, such as temperature or weight, for use in machine learning models.\u003c/p\u003e\n"],["\u003cp\u003eMachine learning practitioners spend significant time on data preparation tasks like cleaning and transformation.\u003c/p\u003e\n"],["\u003cp\u003eThe module covers techniques like feature scaling, outlier detection, and binning to improve data quality for model training.\u003c/p\u003e\n"],["\u003cp\u003eLearners should have a basic understanding of machine learning concepts before starting this module.\u003c/p\u003e\n"],["\u003cp\u003eCategorical data, like postal codes, will be addressed in a separate module due to its distinct characteristics and handling requirements.\u003c/p\u003e\n"]]],[],null,["| **Estimated module length:** 85 minutes\n| **Learning objectives**\n|\n| - Understand feature vectors.\n| - Explore your dataset's potential features visually and mathematically.\n| - Identify outliers.\n| - Understand four different techniques to normalize numerical data.\n| - Understand binning and develop strategies for binning numerical data.\n| - Understand the characteristics of good continuous numerical features.\n| **Prerequisites:**\n|\n| This module assumes you are familiar with the concepts covered in the\n| following module:\n|\n| - [Introduction to Machine Learning](/machine-learning/intro-to-ml)\n\nML practitioners spend far more time evaluating, cleaning, and transforming\ndata than building models.\nData is so important that this course devotes three entire units to the topic:\n\n- Working with numerical data (this unit)\n- [Working with categorical data](/machine-learning/crash-course/categorical-data)\n- [Datasets, generalization, and overfitting](/machine-learning/crash-course/overfitting)\n\nThis unit focuses on\n[**numerical data**](/machine-learning/glossary#numerical-data),\nmeaning integers or floating-point values\nthat behave like numbers. That is, they are additive, countable, ordered,\nand so on. The next unit focuses on\n[**categorical data**](/machine-learning/glossary#categorical-data), which can\ninclude numbers that behave like categories. The third unit focuses on how to\nprepare your data to ensure high-quality results when training and evaluating\nyour model.\n\nExamples of numerical data include:\n\n- Temperature\n- Weight\n- The number of deer wintering in a nature preserve\n\nIn contrast, US postal codes, despite\nbeing five-digit or nine-digit numbers, don't behave like numbers or represent\nmathematical relationships. Postal code 40004 (in Nelson County, Kentucky) is\nnot twice the quantity of postal code 20002 (in Washington, D.C.). These numbers\nrepresent categories, specifically geographic areas, and are considered\ncategorical data.\n| **Key terms:**\n|\n| - [Categorical data](/machine-learning/glossary#categorical-data)\n- [Numerical data](/machine-learning/glossary#numerical-data) \n[Help Center](https://support.google.com/machinelearningeducation)"]]
