A Discrete Wavelet Based Feature Extraction and Hybrid Classification Technique for Microarray Data Analysis

Algorithm 1

Feature selection method by DWT using scaling coefficient and detailed coefficient.

Input: dataset

Output: A subset (selected features by filter method) of the dataset

Method:

(1) Decomposition Step (𝐶: array[1,…,𝑛] of reals)

(2) for 𝑖←1 to 𝑛/2;

(3) 𝐶^′[𝑖] ← (𝐶[2𝑖−1] + 𝐶[2𝑖])/sqrt(2); //scaling coefficient

(4) 𝐶^′[𝑛/2+𝑖] ← (𝐶[2𝑖−1] − 𝐶[2𝑖])/sqrt(2); //detail coefficient

(5) end for

(6) 𝐶←𝐶^′

(7) Threshold application

(8) if 𝐶[𝑖] <= 0.5 then 𝐶[𝑖] = 0; // threshold value = 0.5

(9) else 𝐶[𝑖] = 𝐶[𝑖];

(10) Approximation value using the inverse function

Wavelet inverse 𝑤 = new wavelet(“scaling coef”, “detail coef”);

(11) Feature selection using the filter method

𝑡-test; //rank of 𝑡-test