and decision making support (Kingsford and Salzberg, 2008). In a decision tree, each internal node denotes a test on a variable, and each branch stands for an outcome of the test. Leaf nodes represent an outcome, and the uppermost node in a tree is the root node. The tree can be easily transformed to rules and integrated into computer applications. Decision-tree models can depict classifications based on a set of variables. The ability to create an understandable representation of a classified dataset make decision trees one of the most frequently used data mining techniques (Han and Kamber, 2001). In healthcare, they are commonly used to store and manipulate knowledge, to interpret information, calculate risk, predict outcome, or assess. In operations research, they are often used to help identify the strategy most likely to reach a goal. However, decision trees are data intensive when modeling complex concepts. Large decision trees may be too complex to understand. They are not suitable for continuous-value data that are not easily discretized or for datasets where there are missing data. Also the restriction to a particular tree or rule representation can limit the functionality and approximation power of the model. Decision trees use likelihood-based model evaluation methods, combined with search methods for growing and pruning tree structures. Figure 5.12 The rules acquired from data mining in cases that were not recommended to receive PMRT but against the guideline have received PMRT Razavi et al. (2008) Some breast cancer patients require or choose the surgical procedure of mastectomy, which is complete surgical removal of one or both breasts. To reduce recurrence of cancer and improve overall survival, radiotherapy after mastectomy, or postmastectomy radiotherapy (PMRT) of the chest wall and the regional lymph nodes, is advised. However, the guideline for administering PMRT is not always followed. To maintain the quality of care, Razavi and colleagues data-mined a dataset from a breast cancer registry, identified and extracted patterns of non-compliance (Razavi et al., 2008). By analyzing the group of patients who should not have received PMRT treatment according to the guideline, but received it anyway, the authors constructed a decision tree with seventeen nodes and nine leaves as shown in Figure 5.11. Three of the leaves show non-compliant cases, i.e. cases that received PMRT treatment against the guideline. A set of rules was deduced based on the decision tree as shown in Figure 5.12.
Test Yourself 3.8 A decision tree is a decision support tool that uses a tree-like graph of decisions and their possible consequences. It is a type of a type of rule-based method. True False Key Definitions and Terminology In mathematics, an algorithm is a finite set of elementary instructions that calculate a function or provide result.
You've reached the end of your free preview.
Want to read all 153 pages?
- Spring '19
- Data Mining