You are on page 1of 8

Name: Wafaa Basil Abed Al-Lateef

ID: 220147156
Data Set : Aggregation

1- Apply 3 different clustering algorithms assume the last attribute


as label for computing clustering performant.

classification algorithms: (KMeans , DBSCAN ,K-Medoids).


After testing several possibilities for k
In the end, the best algorithms for cluster these data:
DBSCAN(epsilon=1.5).
2- Apply 3 different classification algorithms.
classification algorithms: (Naive Bayes , DecisionTree ,K-NN).
Sample size =350

Model Accuracy
Naive Bayes 98%
Decision Tree 100%
K-NN , k=3 100%
K-NN , k=7 99.71%
K-NN, k=10 99.71%
K-NN, k=15 99.71%
K-NN, k=20 99.71%
After testing several possibilities for k in K-NN
In the end, the best algorithms for classification these data:
Decision Tree.

You might also like