Incremental Optimization Mechanism for Constructing a Decision Tree in Data Stream Mining

Pseudocode 2

Pseudocode of testing approach.

PROCEDURE:  traverseHT(𝑆, 𝐻𝑇, ℱ)

(1) Sort 𝑆 from the root to a leaf by  HT. Update OCD in each node: 𝑛_𝑖𝑗𝑘(𝑙)++

(2) Switch (ℱ)

(3) Case ℱ^MC: predict the class 𝑦^′_𝑘 with max 𝑛_𝑖𝑗𝑘(𝑙)

(4) Case ℱ^NB: predict the class 𝑦^′_𝑘 with max NB prob.

(5) Case ℱ^WNB: predict the class 𝑦^′_𝑘 with max WNB prob.

(6) IF 𝑦^′_𝑘 equals to the actual class label in 𝑆, THEN 𝐶_𝑇++

(7) ELSE 𝐶_𝐹++

(8)  Δ𝐶=𝐶_𝑇−𝐶_𝐹

(9) Return Δ𝐶