Decision trees are tools used in Machine Learning mainly to perform classification tasks. Samples are submitted to a test in each node of the tree and guided through the tree based on the result. Decision trees can also be used to perform clustering, with a few adjustments: On one hand, new split criteria must be discovered to benefit from the samples features instead of the labels to construct the tree, and on the other hand the nodes referring to the same clusters must be merged. The goal of this thesis is to propose solutions on those two aspects. Unfortunately, when the clustering with decision trees is compared to the existing clustering techniques, the results are poor, with higher misclassification rates on the datasets representing handwritten digits.
- Machine Learning
- Agglomeration
- Decision Tree
- Split Criterion
- Clustering
Clustering with Decision Trees: Divisive and Agglomerative Approach
Castin, L. (Author). 29 Aug 2017
Student thesis: Master types › Master in Computer science