[[["เข้าใจง่าย","easyToUnderstand","thumb-up"],["แก้ปัญหาของฉันได้","solvedMyProblem","thumb-up"],["อื่นๆ","otherUp","thumb-up"]],[["ไม่มีข้อมูลที่ฉันต้องการ","missingTheInformationINeed","thumb-down"],["ซับซ้อนเกินไป/มีหลายขั้นตอนมากเกินไป","tooComplicatedTooManySteps","thumb-down"],["ล้าสมัย","outOfDate","thumb-down"],["ปัญหาเกี่ยวกับการแปล","translationIssue","thumb-down"],["ตัวอย่าง/ปัญหาเกี่ยวกับโค้ด","samplesCodeIssue","thumb-down"],["อื่นๆ","otherDown","thumb-down"]],["อัปเดตล่าสุด 2025-01-03 UTC"],[[["\u003cp\u003eModels ingest data through floating-point arrays called feature vectors, which are derived from dataset features.\u003c/p\u003e\n"],["\u003cp\u003eFeature vectors often utilize processed or transformed values instead of raw dataset values to enhance model learning.\u003c/p\u003e\n"],["\u003cp\u003eFeature engineering is the crucial process of converting raw data into suitable representations for the model, encompassing techniques like normalization and binning.\u003c/p\u003e\n"],["\u003cp\u003eNon-numerical data like strings must be converted into numerical values for use in feature vectors, a key aspect of feature engineering.\u003c/p\u003e\n"]]],[],null,["Until now, we've given you the impression that a model acts directly on the\nrows of a dataset; however, models actually ingest data somewhat differently.\n\nFor example, suppose a dataset provides five columns, but only two of those\ncolumns (`b` and `d`) are features in the model. When processing\nthe example in row 3, does the model simply grab the contents of the\nhighlighted two cells (3b and 3d) as follows?\n**Figure 1.** Not exactly how a model gets its examples.\n\nIn fact, the model actually ingests an array of floating-point values called a\n[**feature vector**](/machine-learning/glossary#feature-vector). You can think\nof a feature vector as the floating-point values comprising one example.\n**Figure 2.** Closer to the truth, but not realistic.\n\nHowever, feature vectors seldom use the dataset's *raw values*.\nInstead, you must typically process the dataset's values into representations\nthat your model can better learn from. So, a more realistic\nfeature vector might look something like this:\n**Figure 3.** A more realistic feature vector.\n\nWouldn't a model produce better predictions by training from the\n*actual* values in the dataset than from *altered* values?\nSurprisingly, the answer is no.\n\nYou must determine the best way to represent raw dataset values as trainable\nvalues in the feature vector.\nThis process is called\n[**feature engineering**](/machine-learning/glossary#feature-engineering),\nand it is a vital part of machine learning.\nThe most common feature engineering techniques are:\n\n- [**Normalization**](/machine-learning/glossary#normalization): Converting numerical values into a standard range.\n- [**Binning**](/machine-learning/glossary#binning) (also referred to as [**bucketing**](/machine-learning/glossary#bucketing)): Converting numerical values into buckets of ranges.\n\nThis unit covers normalizing and binning. The next unit,\n[Working with categorical data](/machine-learning/crash-course/categorical-data),\ncovers other forms of\n[**preprocessing**](/machine-learning/glossary#preprocessing), such as\nconverting non-numerical data, like strings, to floating point values.\n\nEvery value in a feature vector must be a floating-point value. However, many\nfeatures are naturally strings or other non-numerical values. Consequently,\na large part of feature engineering is representing non-numerical values as\nnumerical values. You'll see a lot of this in later modules.\n| **Key terms:**\n|\n| - [Binning](/machine-learning/glossary#binning)\n| - [Bucketing](/machine-learning/glossary#bucketing)\n| - [Feature engineering](/machine-learning/glossary#feature_engineering)\n| - [Feature vector](/machine-learning/glossary#feature_vector)\n| - [Normalization](/machine-learning/glossary#normalization)\n- [Preprocessing](/machine-learning/glossary#preprocessing) \n[Help Center](https://support.google.com/machinelearningeducation)"]]
