Разберись в Data Science. Как освоить науку о данных и научиться думать как эксперт. Джордан Голдмейер

k-ближайших соседей можно использовать для предсказания не только классов, но и чисел. Эти так называемые задачи регрессии мы рассмотрим далее в книге.

      7

      Эта идея обсуждается в чрезвычайно полезной книге Г. Уилсона «Teaching tech together» (CRC Press, 2019).

      8

      Надежная стратегия работы с данными способна смягчить эти проблемы. Разумеется, важным компонентом любой подобной стратегии является решение значимых проблем, и именно на этом мы сосредоточим внимание в этой главе. Если вы хотите узнать больше о высокоуровневой стратегии работы с данными, обратитесь к книге Jagare, U. Data science strategy for dummies. (John Wiley & Sons, 2019).

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