Data Mining Lab

Experiment 7: K Means algorithm

Consider a dataset “bbn.csv” and apply k-means on it. Data link: https://bit.ly/2Hbw2as .

Note:
1) Consider the number of clusters as 2.
2) You can make clusters based on b.tech .
3) Plot the clusters using plot function and replace the points in the graph with their placement status.

Step1: Read the dataset

m<-read.csv("C:/Users/pradeep/OneDrive/datasets/bbn.csv")
head(m)

##   Gender SSC.Percentage inter.Diploma.perc B.Tech.perc Backlogs
## 1      M          87.30               65.3       40.00       18
## 2      F          89.00               92.4       71.45        0
## 3      F          67.00               68.0       45.26       13
## 4      M          71.00               70.4       36.47       17
## 5      M          67.00               65.5       42.52       17
## 6      M          81.26               68.0       62.20        6
##   registered_for_.training Placement.status
## 1                       No       Not placed
## 2                      yes           placed
## 3                      yes       Not placed
## 4                      yes       Not placed
## 5                      yes       Not placed
## 6                      yes       Not placed

Step 2: Apply K-means algorithm.

clust<-kmeans(m$B.Tech.perc,2) # Number of clusters desired are 2.
print(clust)

## K-means clustering with 2 clusters of sizes 67, 50
## 
## Cluster means:
##       [,1]
## 1 71.57092
## 2 51.38185
## 
## Clustering vector:
##   [1] 2 1 2 2 2 1 2 2 2 2 1 2 2 1 2 1 1 1 1 1 1 1 1 1 2 2 2 2 1 1 1 2 1 2 1
##  [36] 1 1 1 2 1 1 2 2 1 1 1 2 1 2 1 1 2 2 1 2 1 2 2 1 1 2 1 2 2 1 1 1 1 2 2
##  [71] 1 2 1 1 1 2 2 1 1 2 1 1 2 1 1 1 2 2 1 1 2 1 1 2 1 1 1 1 1 1 1 1 1 2 1
## [106] 2 2 2 2 2 2 2 2 1 1 1 1
## 
## Within cluster sum of squares by cluster:
## [1] 2189.909 3023.810
##  (between_SS / total_SS =  69.1 %)
## 
## Available components:
## 
## [1] "cluster"      "centers"      "totss"        "withinss"    
## [5] "tot.withinss" "betweenss"    "size"         "iter"        
## [9] "ifault"

3) Plot the clusters.

plot(m$B.Tech.perc,type = "n") # type=n is for no plotting (default is points type="p")

# in the text function, we specify what to plot. 
#"x=m$B.Tech.perc" tells what to plot.
#"labels = m$Placement.status" tells how the points are represented.
#"col = clust$cluster" tells what criteria to follow for coloring

text(x=m$B.Tech.perc,labels = m$Placement.status, col = clust$cluster)