Professional Documents
Culture Documents
QUICK CLUSTER funused0 funused1 funused2 funused3 funused4 funused5 funused6 funused7 funused8 funused9 /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(10) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /PRINT INITIAL.
Initial Cluster Centers Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 1 1 1 1 1 1 1 2 Iteration History(a) Change in Cluster Centers Iteration 1 2 3 4 5 6 7 8 9 10 1 .906 .028 .023 .000 .000 .046 .097 .030 .075 2 1.112 .081 .062 .041 .093 .044 .022 .040 .036 3 .800 .552 .247 .225 .340 .220 .273 .086 .155 2 1 1 2 2 2 2 2 2 2 2 3 2 2 1 2 2 2 1 2 1 1
.021 .019 .060 a Iterations stopped because the maximum number of iterations was performed. Iterations failed to converge. The maximum absolute coordinate change for any center is .034. The current iteration is 10. The minimum distance between initial centers is 2.449.
Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 2 1 1 1 2 1 1 2 2 1 1 2 2 2 1 2 1 2 2 3 1 2 1 2 2 1 2 2 1 2
Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 86.000 73.000 47.000 206.000 .000
QUICK CLUSTER funused0 funused1 funused2 funused3 funused4 funused5 funused6 funused7 funused8 funused9 /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(20) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /PRINT INITIAL.
Quick Cluster
Initial Cluster Centers Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 1 1 1 1 1 1 1 2 2 1 1 2 2 2 2 2 2 2 2 3 2 2 1 2 2 2 1 2 1 1
Iteration History(a) Change in Cluster Centers Iteration 1 2 3 4 5 6 7 8 9 10 11 1 .906 .028 .023 .000 .000 .046 .097 .030 .075 .021 2 1.112 .081 .062 .041 .093 .044 .022 .040 .036 .019 3 .800 .552 .247 .225 .340 .220 .273 .086 .155 .060
.000 .000 .000 a Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 11. The minimum distance between initial centers is 2.449. Final Cluster Centers Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 2 1 1 1 2 1 1 2 2 1 1 2 2 2 1 2 1 2 2 3 1 2 1 2 2 1 2 2 1 2
Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 86.000 73.000 47.000 206.000 .000
QUICK CLUSTER funused0 funused1 funused2 funused3 funused4 funused5 funused6 funused7 funused8 funused9 /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(20) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /SAVE CLUSTER /PRINT INITIAL.
qqqqqq
Initial Cluster Centers Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 1 1 1 1 1 1 1 2 Iteration History(a) Change in Cluster Centers Iteration 1 2 3 4 5 6 7 8 9 10 11 1 .906 .028 .023 .000 .000 .046 .097 .030 .075 .021 2 1.112 .081 .062 .041 .093 .044 .022 .040 .036 .019 3 .800 .552 .247 .225 .340 .220 .273 .086 .155 .060 2 1 1 2 2 2 2 2 2 2 2 3 2 2 1 2 2 2 1 2 1 1
.000 .000 .000 a Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 11. The minimum distance between initial centers is 2.449. Final Cluster Centers Cluster 1 SMS Alarm Camera Scheduler Music / Radio Playback Games Internet Time and Date Download Other 1 1 2 1 1 1 2 1 1 2 2 1 1 2 2 2 1 2 1 2 2 3 1 2 1 2 2 1 2 2 1 2
Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 86.000 73.000 47.000 206.000 .000
USE ALL. COMPUTE filter_$=(QCL_1 = 3). VARIABLE LABEL filter_$ 'QCL_1 = 3 (FILTER)'. VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'. FORMAT filter_$ (f1.0). FILTER BY filter_$. EXECUTE . FREQUENCIES VARIABLES=gender educatn service contype chrgfreq paymode /ORDER= ANALYSIS .
Frequencies
Statistics Name of current service provider 47 0
Valid Missing
Gender of respondent 47 0
Level of education 47 0
Connection Type 47 0
Mode of payment 47 0
Frequency Table
Gender of respondent Cumulative Percent 85.1 100.0
Valid
Frequency 40 7 47
Valid
Frequency 1 18 28 47
Valid
Frequency 12 8 13 3 11 47
Valid
Frequency 42 5 47
Valid
Frequency 3 41 2 1 47
Valid
Frequency 45 1 1 47
QUICK CLUSTER mntspend billsms billothr billtalk billfix /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(10) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /PRINT INITIAL.
Quick Cluster
Initial Cluster Centers Cluster 1 Monthly expenditure on phone 1000.00 2 2000.00 3 99.00
SMS bill Other charges Voice calls bill Fixed component of bill
Iteration History(a) Change in Cluster Centers 1 2 3 252.234 .000 216.653 .000 .000 .000 a Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 2. The minimum distance between initial centers is 901.347. Final Cluster Centers Cluster 1 Monthly expenditure on phone SMS bill Other charges Voice calls bill Fixed component of bill 750.00 35.00 2.50 43.75 31.25 2 2000.00 40.00 .00 60.00 70.00 3 313.24 27.26 3.95 43.10 48.64 Iteration 1 2
Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 4.000 1.000 42.000 47.000 .000
// Begin here Clusters based on monthly expenditure on phone is made using K-means method. The following clusters are obtained. Cluster 2 has the maximum expenditure.
FILTER OFF. USE ALL. EXECUTE . QUICK CLUSTER mntspend billsms billothr billtalk billfix /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(10) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /PRINT INITIAL.
Quick Cluster
[DataSet1] C:\Users\Dell\Downloads\Cell_Inter.sav
Initial Cluster Centers
Cluster 1 Monthly expenditure on phone SMS bill Other charges Voice calls bill Fixed component of bill 1000.00 25.00 .00 75.00 25.00 2 2000.00 40.00 .00 60.00 70.00 3 99.00 10.00 .00 10.00 80.00
Iteration History(a) Change in Cluster Centers 1 2 3 267.544 .000 225.437 .000 .000 .000 a Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 2. The minimum distance between initial centers is 905.139. Final Cluster Centers Cluster 1 Monthly expenditure on phone SMS bill Other charges Voice calls bill Fixed component of bill 734.69 27.85 3.08 46.54 44.08 2 2000.00 40.00 .00 60.00 70.00 3 318.14 26.65 5.77 48.54 48.29 Iteration 1 2
It reduced the distance between the clusters centers after the maximum absolute coordinate change for any center is obtained as .000. Now the three clusters with the changed mean value are obtained as above. Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 13.000 1.000 192.000 206.000 .000
VARIABLES=mntspend /COMPARE VARIABLE/PLOT=BOXPLOT/STATISTICS=NONE/NOTOTAL /MISSING=LISTWISE . It could be seen that some cases are outliers and extremes, to make the consistent clusters these were removed. A boxplot was made to see this.
Explore
Case Processing Summary Cases Missing
Valid
Total
Percent 100.0%
N 0
Percent .0%
N 206
Percent 100.0%
2000.00
39
1500.00
35 77 500.00
104 134 81 89
151 2
USE ALL. COMPUTE filter_$=(mntspend < 600). VARIABLE LABEL filter_$ 'mntspend < 600 (FILTER)'. VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'. FORMAT filter_$ (f1.0). FILTER BY filter_$. EXECUTE . QUICK CLUSTER mntspend billsms billothr billtalk billfix /MISSING=LISTWISE /CRITERIA= CLUSTER(3) MXITER(10) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /PRINT INITIAL. The data was filtered and the filter criterion was the monthly expenditure less than Rs. 600 was kept and the expenditure above this was kept aside.
Quick Cluster
Initial Cluster Centers Cluster 1 2 3
Monthly expenditure on phone SMS bill Other charges Voice calls bill Fixed component of bill
Iteration History(a) Change in Cluster Centers Iteration 1 2 3 1 71.530 12.762 2 72.482 7.535 3 97.952 19.970
.000 .000 .000 a Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 3. The minimum distance between initial centers is 250.450. Final Cluster Centers Cluster 1 Monthly expenditure on phone SMS bill Other charges Voice calls bill Fixed component of bill 339.50 27.66 6.41 45.69 46.34 2 488.56 31.48 5.07 49.63 43.63 3 208.52 22.57 4.39 53.61 53.41
The above data tables were obtained on again taking out K-Means clusters. Out of 206 cases, 9 were kept aside. Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 116.000 27.000 54.000 197.000 .000
sort cases by QCL_2 (a) . SORT CASES BY QCL_2 . SPLIT FILE LAYERED BY QCL_2 . . The 9 extremes were manually kept in cluster 2 as it was the cluster of high expenditure. And a new variable of clusters was added in the table. We pick the cluster 2, from the above table it is clear that this group has the highest expenditure on the phone. CROSSTABS /TABLES=QCL_2 BY contype /FORMAT= AVALUE TABLES /CELLS= COUNT ROW COLUMN /COUNT ROUND CELL .
Crosstabs
[DataSet1] C:\Users\Dell\Downloads\Cell_Inter.sav
Case Processing Summary Cases Valid N Cluster Number of Case * Connection Type 197 Percent 95.6% N 9 Missing Percent 4.4% N 206 Total Percent 100.0%
Cluster Number of Case * Connection Type Crosstabulation Connection Type Cluster Number of Case 1 Count % within Cluster Number of Case % within Connection Type 2 Count % within Cluster Number of Case % within Connection Type 3 Count % within Cluster Number of Case % within Connection Type Total Count % within Cluster Number of Case % within Connection Type Prepaid 110 94.8% 61.1% 23 85.2% 12.8% 47 87.0% 26.1% 180 91.4% 100.0% Post Paid 6 5.2% 35.3% 4 14.8% 23.5% 7 13.0% 41.2% 17 8.6% 100.0% Total Prepaid 116 100.0% 58.9% 27 100.0% 13.7% 54 100.0% 27.4% 197 100.0% 100.0%
It is clear from the above table that the more people falling in cluster 2 have the Post paid connection as compared to the other type of connection. Here by, we make an assumption that people having high expenditure on phone usually have the post paid connection.
CROSSTABS /TABLES=contype BY funused5 /FORMAT= AVALUE TABLES /CELLS= COUNT ROW COLUMN /COUNT ROUND CELL .
Crosstabs
[DataSet1] C:\Users\Dell\Downloads\Cell_Inter.sav
Case Processing Summary Cases Valid N Connection Type * Games 206 Percent 100.0% N 0 Missing Percent .0% N 206 Total Percent 100.0%
Connection Type * Games Crosstabulation Games Connection Type Prepaid Count % within Connection Type % within Games Post Paid Count % within Connection Type % within Games Total Count % within Connection Type % within Games Yes 158 85.9% 88.8% 20 90.9% 11.2% 178 86.4% 100.0% No 26 14.1% 92.9% 2 9.1% 7.1% 28 13.6% 100.0% Total Yes 184 100.0% 89.3% 22 100.0% 10.7% 206 100.0% 100.0%
The above cross tab shows that majority of post paid subscribers play games as compared to the pre paid connection. Hence, it could be deducted that the cluster 2 people have high phone expenditure, prefer post paid plans and play games.
CROSSTABS /TABLES=QCL_2 BY funused0 /FORMAT= AVALUE TABLES /CELLS= COUNT ROW COLUMN /COUNT ROUND CELL .
Crosstabs
[DataSet1] C:\Users\Dell\Downloads\Cell_Inter.sav
Case Processing Summary Cases Valid N Cluster Number of Case * SMS 197 Percent 95.6% N 9 Missing Percent 4.4% N 206 Total Percent 100.0%
Cluster Number of Case * SMS Crosstabulation SMS Cluster Number of Case 1 Count % within Cluster Number of Case % within SMS 2 Count % within Cluster Number of Case % within SMS 3 Count % within Cluster Number of Case % within SMS Total Count % within Cluster Number of Case Yes 103 88.8% 57.5% 26 96.3% 14.5% 50 92.6% 27.9% 179 90.9% No 13 11.2% 72.2% 1 3.7% 5.6% 4 7.4% 22.2% 18 9.1% Total Yes 116 100.0% 58.9% 27 100.0% 13.7% 54 100.0% 27.4% 197 100.0%
% within SMS
100.0%
100.0%
100.0%
The above cross tab shows that the cluster 2 people (96.3%) do use SMS service, which is significantly higher than that of the pre paid users. Uptil here, we get that CLUSTER 2 people spend the highest on the phone, like games and SMS services and have postpaid plan.