Методы обучения без учителя

Уровень 0 · материалов: 6

В кластер входят документы, посвященные алгоритмам и теоретическим основам обучения без учителя, включая кластеризацию и снижение размерности данных.

Общие признаки: обучение без учителя, кластеризация данных, снижение размерности, алгоритмы PCA, алгоритмы группировки данных

Группа выше: Методы обучения и оптимизации моделей

Материалы

• #25135 — Unsupervised Learning: PCA and Clustering Guide — Представитель этого кластера

  Открытый курс машинного обучения. Тема 7. Обучение без учителя: PCA и кластеризация — libfun

  Смысл: The main idea of the text is to provide a theoretical and practical introduction to unsupervised learning, specifically teaching how to reduce data dimensionality via PCA and how to group similar data points using various clustering algorithms.

  An educational guide explaining the theory and Python implementation of PCA for dimensionality reduction and multiple clustering algorithms for grouping unlabeled data.

• #9026 — Overview of Data Clustering Algorithms

  Обзор алгоритмов кластеризации данных — andreycha

  Смысл: The main idea is to provide a structured educational overview of different data clustering methodologies, explaining their mathematical foundations, classifications, and practical trade-offs to help researchers choose the appropriate algorithm for their specific data analysis task.

  A technical guide explaining the principles, distance metrics, and various types of data clustering algorithms, including hierarchical, flat, fuzzy, and graph-based methods.

• #30386 — Introduction to Unsupervised Learning and Kohonen Maps

  Unsupervised learning или «пойди туда, не знаю куда, найди то, не знаю что» — stepan_ovchinnikov

  Смысл: The main idea is to introduce the concept of unsupervised learning through the lens of Kohonen's Self-Organizing Maps, explaining how high-dimensional, unlabeled data can be transformed into a structured, visual map to uncover hidden patterns.

  An overview of unsupervised learning focusing on how Kohonen Maps organize high-dimensional data into clusters without predefined labels.

• #2877 — Understanding Principal Component Analysis (PCA) via Simple Example

  Как работает метод главных компонент (PCA) на простом примере — Lexxo

  Смысл: The text aims to explain the intuitive and mathematical workings of Principal Component Analysis (PCA) using a simple numerical example. It demonstrates how to reduce the number of features in a dataset while preserving the maximum amount of variance, effectively simplifying data models without significant loss of information.

  A comprehensive guide explaining the intuition and mathematics of PCA, featuring a step-by-step Python implementation and comparison with scikit-learn.

• #31497 — Understanding the DBSCAN Clustering Algorithm

  Интересные алгоритмы кластеризации, часть вторая: DBSCAN — Siarshai

  Смысл: The main idea is to explain the mechanics, mathematical foundation, and practical utility of the DBSCAN clustering algorithm, emphasizing its strength in detecting clusters of arbitrary shapes and its resilience to noise compared to centroid-based methods.

  A detailed technical explanation of the DBSCAN clustering algorithm, covering its intuitive logic, mathematical framework, parameter tuning, and practical advantages in handling noise and complex cluster shapes.

• #22784 — Clustering Algorithms K-means and C-means

  Кластеризация: алгоритмы k-means и c-means — sashaeve

  Смысл: The main idea is to explain the concept of data clustering and compare the k-means and c-means algorithms, highlighting that while k-means is simpler, c-means is more flexible because it allows for probabilistic membership of objects in clusters.

  An overview of clustering in Data Mining, comparing the rigid k-means algorithm with the probabilistic c-means approach and explaining various distance metrics.