Scribd Logo
Upload

Welcome to Scribd!

  • 6 views

    Clustering

    Uploaded by

    ExpoMed ExpoMed
    Clustering is an unsupervised machine learning technique used to group unlabeled data points into meaningful clusters. There are several types of clustering algorithms, including k-means clustering, DBSCAN clustering, and self-organizing maps (SOM). K-means clustering partitions data into k groups by finding cluster centroids. DBSCAN clustering identifies core data points that form dense clusters based on a distance epsilon. SOM creates a visual map to represent similarities between data by adjusting neuron weights.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Clustering

    Uploaded by

    ExpoMed ExpoMed
    6 views26 pages
    Clustering is an unsupervised machine learning technique used to group unlabeled data points into meaningful clusters. There are several types of clustering algorithms, including k-means clustering, DBSCAN clustering, and self-organizing maps (SOM). K-means clustering partitions data into k groups by finding cluster centroids. DBSCAN clustering identifies core data points that form dense clusters based on a distance epsilon. SOM creates a visual map to represent similarities between data by adjusting neuron weights.

    Original Title

    07. Clustering

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Clustering is an unsupervised machine learning technique used to group unlabeled data points into meaningful clusters. There are several types of clustering algorithms, including k-means clustering, DBSCAN clustering, and self-organizing maps (SOM). K-means clustering partitions data into k groups by finding cluster centroids. DBSCAN clustering identifies core data points that form dense clusters based on a distance epsilon. SOM creates a visual map to represent similarities between data by adjusting neuron weights.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    6 views26 pages

    Clustering

    Uploaded by

    ExpoMed ExpoMed
    Clustering is an unsupervised machine learning technique used to group unlabeled data points into meaningful clusters. There are several types of clustering algorithms, including k-means clustering, DBSCAN clustering, and self-organizing maps (SOM). K-means clustering partitions data into k groups by finding cluster centroids. DBSCAN clustering identifies core data points that form dense clusters based on a distance epsilon. SOM creates a visual map to represent similarities between data by adjusting neuron weights.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 26

    7.

    Clustering
    Clustering

    Clustering is the process of finding meaningful groups in data.

    For example, customers of a company can be grouped based on the purchase


    behavior. In recent years, clustering has even found its use in political elections

    Clustering to describe the data

    Clustering for pre-processing


    Types of clustering
    1. Exclusive or strict partitioning clusters
    2. Overlapping clusters
    3. Hierarchical clusters
    4. Fuzzy or probabilistic clusters

    Based Algorithmic approach

    1. Prototype-based clustering
    2. Density clustering
    3. Hierarchical clustering
    4. Model-based clustering
    k-Means Clustering
    k-Means
    k-means clustering is a prototype-based clustering method where the data set
    is divided into k clusters.

    Objective: find a prototype data point for each cluster; all the data points are
    then assigned to the nearest prototype, which then forms a cluster
    k Partitions

    k-means algorithm divides the data space


    into k partitions or boundaries, where the
    centroid in each partition is the prototype of
    the clusters

    Voronoi partition. (“Euclidean Voronoi Diagram” by Raincomplex – personal work. Licensed under Creative
    Commons Zero, Public Domain Dedication via Wikimedia Commons
    k Partitions

    k-means algorithm divides


    the data space into k
    partitions or boundaries,
    where the centroid in each
    partition is the prototype of
    the clusters
    Step 1: Initiate Centroids
    Step 2: Assign Data Points
    Step 3: Calculate New Centroids

    Minimizing the sum of


    squared errors (SSE)
    Step 4: Repeat Assignment and Calculate
    New Centroids
    Step 5: Termination

    No further change in assignment of data points happens or, in other words, no


    significant change in centroids are noted.

    Evaluation of Clusters

    - Minimize total SSE


    - Davies-Bouldin index
    DBSCAN CLUSTERING
    Density in the dataset
    Density of a data point

    The number of points within a circular space


    with radius ε (epsilon) around a data point A
    is six
    Step 1: Defining Epsilon and MinPoints

    The number of data points inside the space is defined by radius ε . If


    MinPoints is defined as 5, the space ε surrounding data point A is considered
    a high-density region.
    Step 2: Classification of Data Points
    Step 3: Clustering

    Groups of core points form distinct clusters. If two core points are within ε of
    each other, then both core points are within the same cluster.

    Optimize the Parameters: ε and a minimum threshold (MinPoints)


    Special Cases: Varying Densities
    SELF-ORGANIZING MAPS
    SOM
    Powerful visual clustering technique

    A neural network. Output is an organized visual matrix. SOM output is a


    two-dimensional grid with data objects placed next to each other based on
    their similarity to one another.
    Step 1: Topology Specification
    Step 2: Initialize Centroids

    The initial centroids are values of random data objects from the data set.

    Step 3: Assignment of Data Objects

    Data objects are selected one by one and assigned to the nearest centroid.
    Step 4: Centroid Update

    Update the data values of the nearest centroid of the data object, proportional
    to the difference between the centroid and the data object.
    Step 4: Centroid Update
    Step 5: Termination

    Until no significant centroid updates take place in each run.

    Step 6: Mapping a New Data Object

    Based on proximity to the centroids.

    You might also like