Supervised Learning

Supervised learning is a machine learning approach in which every training example pairs input features with the correct output, or label. The model learns the mapping and is then used to predict labels for new inputs.

Why it matters

Most practical ML falls here: spam detection, price prediction, fraud flagging, and image classification. Its main constraint is labeling, since accurate labels often need human time. Good results depend on having enough labeled examples that represent the cases you will meet later.

Example

Two common problem types:

Classification: email -> "spam" or "not spam"
Regression:     house size, location -> price (a number)

Performance is measured on a held-out test set, using metrics such as accuracy, precision, recall, or mean error.

Common mistakes

• Using accuracy alone on imbalanced data; a model predicting "not fraud" every time can be 99% accurate and useless.
• Labels that are inconsistent across annotators.
• Evaluating on data the model has already seen, hiding overfitting.

Related

• Unsupervised learning
• Training data
• Machine learning
• Dataset

यह पेज सामान्य जानकारी है, जिसे हमारी संपादकीय टीम ने लिखा है और दस्तावेज़ीकरण बदलने पर अपडेट किया जाता है। सर्वर और प्रदाता अलग-अलग होते हैं, इसलिए बदलाव पहले सुरक्षित तरीके से जाँचें। देखें: संपादकीय नीति और अस्वीकरण.