Bioinformatics and Medical Applications. Группа авторов

Bioinformatics and Medical Applications - Группа авторов


Скачать книгу
order of individual accuracy).

       • Parallel: All three algorithms are applied in parallel and maximum voting is used.

       • Prob 60 SP: If probability calculated by Naive Bayes is greater than 60% apply serial method else apply parallel.

       • PLS: First parallel then serial is applied for wrong classified records.

       • SKmeans: Combination of Serial along with K means.

       • PKmeans: Combination of Parallel along with K means.

Schematic illustration of the NB confusion matrix. Schematic illustration of the RF confusion matrix.

      Figure 1.9 RF confusion matrix.

      Figure 1.10 DT confusion matrix.

Graph depicts the ROC curve analysis.

      1.4.2 Method

       initialization

       d ← dataset

       a1 ← Naive_Bayes_output ← ApplyNaiveBayes(d)

       a2 ← Decision_tree_output ← ApplyDesisionTree(d)

       a3 ← Random_tree_output ← ApplyRandomForest(d)

       a4 ← K_Means_output ← ApplyKmeans(d)

       winner(0, 1) ← Voting(a1, a2, a3, a4)

       op ← winner_of_max_count(0,1)

       if op ≠ desired_output then

       Probability_calculation of each column with output 0 or 1

image

      end

      For each value in ci

      countci/2

      For k to count

       Add the probability (Find the max column with which probability matches)

       Number of columns selected as ti

       wi ← Weightage of selected columns

       αi ← Append the weightage with the input of data

      Find mean square error with the training and find lowest (MSE) parameter. Calculate the Euclidean distance

image

      Find the minimum distance using this formula.

      If probability of data > 0.5 and MSE < 0.5 and ED < 0.2

      Classify as 1

      else

      Classify as 0

      The following block diagram explains the flow of Algorithm 1.1.

Schematic illustration of the block diagram of the algorithm 1.

      1 1. The ensemble method of the four algorithms (Decision Tree, Random Forest, Naive Bayes, and K Means) is applied by majority voting and classification is obtained on presence or absence of cardiopathy.

      2 2. The wrongly classified records are stored in a separate dataset.

      3 3. The probability of each column with output is calculated and stored. For example, considering age, the probability of heart disease for age greater than 45 is more than otherwise.

      4 4. We calculate those columns for which probability is maximum.

      5 5. Only select these columns for further analysis.

      6 6. Calculate the weights of these columns using formula y = mx + c for linear data using Multiple linear regression.

      7 7. For non-linear data wherein the chances of misclassification are more, more complex functions such as tanh, sigmoid, and relu are used for calculating the weights.

      8 8. Append the weights to the column at the time of classification.

      9 9. Calculate the mean square error and Euclidean distance.

      10 10. Finally, based on probability, mean square error and Euclidean distance, we classify the records as 1 or 0 which indicates presence/absence of heart disease.

      11 11. Hence, accuracy achieved is higher than using the classical ensemble method.

      Hence, our proposed methodology achieves a precision that not only surpasses the individual methods but also overshoots the combination method and the precision achieved thus is quite competitive.


Скачать книгу