Experiment 8: Agglomerative hierarchal clustering

  1. Consider a dataset “hclustdata.csv” and apply agglomerative hierarchal clustering on it. Data link:https://bit.ly/2Hb4Sk4 .
  1. Apply using single link ( Min ) technique.
  2. Apply using complete link (Max) technique.
  3. Apply using group average technique.
    Note: Use Euclidean distance to compute proximity matrix. Check weather 14th entry and 16th entry are similar or not

Step 1: Import the package.

library("cluster")

Step 2: Import the dataset

m<-read.csv("C:/Users/pradeep/OneDrive/datasets/hclustdata.csv")
head(m)
##                    Name Gender SSC.Perc.entage inter.Diploma.perc
## 1       ARIGELA AVINASH      M           87.30               65.3
## 2    BALADARI KEERTHANA      F           89.00               92.4
## 3 BAVIRISETTI PRAVALIKA      F           67.00               68.0
## 4        BODDU SAI BABA      M           71.00               70.4
## 5    BONDAPALLISRINIVAS      M           67.00               65.5
## 6         CH KANAKARAJU      M           81.26               68.0
##   B.Tech.perc Back.logs
## 1       40.00        18
## 2       71.45         0
## 3       45.26        13
## 4       36.47        17
## 5       42.52        17
## 6       62.20         6

Step 3b: Apply Agglomerative hierarchal clustering with group group average

clust3<-agnes(x = m,stand = TRUE,metric = "euclidean",method = "average")
pltree(clust3)

m[c(14,16),] # Check whether 14 and 16 are in same cluster or not
##                 Name Gender SSC.Perc.entage inter.Diploma.perc B.Tech.perc
## 14        EDARA ROJA      F           87.10               88.7       74.96
## 16 GADIPALLI MADHURI      F           85.83               87.0       75.96
##    Back.logs
## 14         0
## 16         0