Scribd Logo
Upload

Welcome to Scribd!

  • 24 views

    K Means

    Uploaded by

    mohamed mohsen
    The document describes performing K-Means clustering on a dataset. It loads the necessary libraries and data. It then determines the optimal number of clusters (K) to use by applying the Elbow method, gap statistic, and Silhouette method. The Elbow method suggests K=5 clusters, the gap statistic K=1 (which is ignored), and the Silhouette method K=8 clusters. It then performs K-Means clustering with K=5 and K=8, plots the results, and calculates the mean of each cluster.

    Copyright:

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

    K Means

    Uploaded by

    mohamed mohsen
    24 views4 pages
    The document describes performing K-Means clustering on a dataset. It loads the necessary libraries and data. It then determines the optimal number of clusters (K) to use by applying the Elbow method, gap statistic, and Silhouette method. The Elbow method suggests K=5 clusters, the gap statistic K=1 (which is ignored), and the Silhouette method K=8 clusters. It then performs K-Means clustering with K=5 and K=8, plots the results, and calculates the mean of each cluster.

    Original Description:

    Original Title

    K-Means

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document describes performing K-Means clustering on a dataset. It loads the necessary libraries and data. It then determines the optimal number of clusters (K) to use by applying the Elbow method, gap statistic, and Silhouette method. The Elbow method suggests K=5 clusters, the gap statistic K=1 (which is ignored), and the Silhouette method K=8 clusters. It then performs K-Means clustering with K=5 and K=8, plots the results, and calculates the mean of each cluster.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    24 views4 pages

    K Means

    Uploaded by

    mohamed mohsen
    The document describes performing K-Means clustering on a dataset. It loads the necessary libraries and data. It then determines the optimal number of clusters (K) to use by applying the Elbow method, gap statistic, and Silhouette method. The Elbow method suggests K=5 clusters, the gap statistic K=1 (which is ignored), and the Silhouette method K=8 clusters. It then performs K-Means clustering with K=5 and K=8, plots the results, and calculates the mean of each cluster.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 4

    K-Means clustering method

    1. install and call need libraries:


    library("readxl")
    library("ggplot2")
    library("dplyr")
    library("ggfortify")
    library("factoextra")
    library("cluster")
    2. read data from data1.xlsx file, then select columns 2,3:
    data <-read_excel('data 1.xlsx')
    df<- select(data,c(2:3))
    3. Determinant the value of K.
    A. Using the Elbow Method:
    # function to compute total within-cluster sum of squares
    fviz_nbclust(df, kmeans, method = "wss", k.max = 24) + theme_minimal() + ggtitle("the Elbow Method")
    Output:

    According to the Elbow Method, the optimal value of K is 5.


    B. Using The gap statistic Method:
    # The gap statistic compares the total within intra-cluster variation for different values of k with their expected values under null
    reference distribution of the data.
    # >>>> The estimate of the optimal clusters will be value that maximize the gap statistic
    gap_stat <- clusGap(df, FUN = kmeans, nstart = 30, K.max = 24, B = 50)
    fviz_gap_stat(gap_stat) + theme_minimal() + ggtitle("fviz_gap_stat: Gap Statistic")
    Output:

    According to the gap statistic Method, the optimal value of K is 1.


    So we will ignore it.
    C. Using The Silhouette Method:
    # Average silhouette method computes the average silhouette of observations for different values of k
    #The optimal number of clusters k is the one that maximize the average silhouette over a range of possible values for k.
    fviz_nbclust(df, kmeans, method = "silhouette", k.max = 24) + theme_minimal() + ggtitle("The Silhouette Plot")
    Output:

    According to the Silhouette Method, the optimal value of K is 8.

    4. K-Means clustering method


    A. K=5:
    #perform k-means clustering with k = 5 clusters
    km_c5 <- kmeans(df, centers = 5, nstart = 25)
    #plot results of final k-means model
    fviz_cluster(km_c5, data = df)
    Output:

    find the mean for each cluster and add new column in data for Cluster.
    #function to find the mean of the variables in each cluster
    aggregate(df, by=list(cluster=km_c5$cluster), mean)
    final_data <- cbind(df, cluster = km_c5$cluster)
    Output:
    Cluster times accel
    1 42.00000 5.12500
    2 10.94545 -2.90000
    3 35.58182 -22.14545
    4 33.22000 53.18000
    5 26.30000 -86.30000
    B. K=8:
    #perform k-means clustering with k = 8 clusters
    km_c8 <- kmeans(df, centers = 8, nstart = 25)
    #plot results of final k-means model
    fviz_cluster(km_c8, data = df)
    Output:

    find the mean for each cluster and add new column in data for Cluster.
    #function to find the mean of the variables in each cluster
    aggregate(df, by=list(cluster=km_c8$cluster), mean)
    final_data <- cbind(df, cluster = km_c8$cluster)
    Output:
    Cluster times accel
    1 32.28571 -18.928571
    2 32.42500 5.012500
    3 53.22857 -1.714286
    4 26.20000 -107.000000
    5 33.73333 73.200000
    6 9.44000 -2.650000
    7 33.00000 44.600000
    8 29.93333 -49.566667
    Consolidated code:
    # Import Set Of Library
    library("readxl")
    library("ggplot2")
    library("dplyr")
    library("ggfortify")
    library("factoextra")
    library("cluster")
    #----------------------------------
    # Load Data From Excel File
    data <-read_excel('data 1.xlsx')
    View(data)
    #-----------------------------------
    # Select Data That Use in Training
    df<- select(data,c(2:3))
    #------------------------------------
    set.seed(1) #it ensures that you get the same result always.
    # The Elbow Method :
    # function to compute total within-cluster sum of squares
    fviz_nbclust(df, kmeans, method = "wss", k.max = 24) + theme_minimal() + ggtitle("the Elbow Method")
    #------------------------------------
    # The gap statistic compares the total within
    # intra-cluster variation for different values of k with their expected values
    # under null reference distribution of the data.
    # >>>> The estimate of the optimal clusters will be value that maximize the gap statistic
    gap_stat <- clusGap(df, FUN = kmeans, nstart = 30, K.max = 24, B = 50)
    fviz_gap_stat(gap_stat) + theme_minimal() + ggtitle("fviz_gap_stat: Gap Statistic")
    #-------------------------------------------------------
    # The Silhouette Method :
    # Average silhouette method computes the average silhouette of observations for different values of k
    # >>>> The optimal number of clusters k is the one that maximize the average silhouette over a range of possible values for k.
    fviz_nbclust(df, kmeans, method = "silhouette", k.max = 24) + theme_minimal() + ggtitle("The Silhouette Plot")
    #-------------------------------------------------------------
    #perform k-means clustering with k = 5 clusters
    km_c5 <- kmeans(df, centers = 5, nstart = 25)
    #plot results of final k-means model
    fviz_cluster(km_c5, data = df)
    #function to find the mean of the variables in each cluster
    aggregate(df, by=list(cluster=km_c5$cluster), mean)
    # We can also append the cluster assignments of each state back to the original dataset:
    #add cluster assigment to original data
    final_data <- cbind(df, cluster = km_c5$cluster)
    #view final data
    final_data
    #--------------------------------------------------------------
    #perform k-means clustering with k = 8 clusters
    km_c8 <- kmeans(df, centers = 8, nstart = 25)
    #plot results of final k-means model
    fviz_cluster(km_c8, data = df)
    #function to find the mean of the variables in each cluster
    aggregate(df, by=list(cluster=km_c8$cluster), mean)
    # We can also append the cluster assignments of each state back to the original dataset:
    #add cluster assigment to original data
    final_data <- cbind(df, cluster = km_c8$cluster)
    #view final data
    final_data

    You might also like