public Algorithms(ITestOutputHelper output)
{
alg01 = new Algorithm01(0.1, 0.8);
var psi = new Dictionary <string, double>();
psi["c1"] = 0.8;
psi["c2"] = 0.6;
alg02 = new Algorithm02(0.1, psi);
this.output = output;
}
static void performAlg01(FuzzyTable table, int indexForParam)
{
Console.WriteLine("performAlg01: " + indexForParam);
int size = 11;
Double[] kriteriaArray = new Double[size];
for (int i = 0; i < size; i++)
{
var beta = 0.0 + 0.1 * i;
var dataSize = 0;
for (int j = 0; j < 200; j++)
{
Algorithm01 alg01;
switch (indexForParam)
{
case 0:
alg01 = new Algorithm01(beta, 0.7);
break;
case 1:
default:
alg01 = new Algorithm01(0, beta);
break;
}
alg01.init(table);
var validation02 = new TenCrossValidation();
var matrix02 = validation02.Validate(10, table, alg01);
if (matrix02 != null)
{
var kriteria = (matrix02.Sensitivity() + matrix02.Specificity()) / 2;
kriteriaArray[i] += kriteria;
dataSize++;
}
}
Console.WriteLine(indexForParam + " CURRENT beta: " + (beta) + " :" + kriteriaArray[i] / dataSize);
}
}
static void Main(string[] args)
{
var table = new FuzzyTable();
try
{ // Open the text file using a stream reader.
using (StreamReader sr = new StreamReader("./data/bupa_fuzzy.json"))
{
String json = sr.ReadToEnd();
dynamic array = JsonConvert.DeserializeObject(json);
for (int i = 0; i < array.attributes.Count - 1; i++)
{
table.addAttribute(array.attributes[i]);
}
table.addClassAttribute(array.attributes[array.attributes.Count - 1], "yes", "no");
table.AddData(array.data);
Console.WriteLine("Started");
var alg = new Algorithm01(0, 0.6);
// var alg = new Algorithm01Modification(0, 0.62, 0.2);
int maxReplications = 30;
int replications = 0;
double TP = 0;
double FP = 0;
double TN = 0;
double FN = 0;
double accuracy = 0;
double sensitivity = 0;
double specificity = 0;
double precision = 0;
for (int i = 0; i < maxReplications; i++)
{
alg.init(table);
var validation01 = new TenCrossValidation();
var matrix = validation01.Validate(10, table, alg);
if (matrix != null)
{
accuracy += matrix.Accuracy();
sensitivity += matrix.Sensitivity();
specificity += matrix.Specificity();
precision += matrix.Precision();
TP += matrix.TruePositivePercent;
FP += matrix.FalsePositivePercent;
TN += matrix.TrueNegativePercent;
FN += matrix.FalseNegativePercent;
replications++;
}
Console.WriteLine(i + " Criteria: " + ((sensitivity / replications) + (specificity / replications)) / 2);
}
Console.WriteLine();
Console.WriteLine("Replications: " + replications);
Console.WriteLine("TP: " + TP / replications);
Console.WriteLine("FP: " + FP / replications);
Console.WriteLine("TN: " + TN / replications);
Console.WriteLine("FN: " + FN / replications);
Console.WriteLine("Accuracy: " + accuracy / replications);
Console.WriteLine("Sensitivity: " + sensitivity / replications);
Console.WriteLine("Specificity: " + specificity / replications);
Console.WriteLine("Precision: " + precision / replications);
Console.WriteLine("Criteria: " + ((sensitivity / replications) + (specificity / replications)) / 2);
return;
}
}
catch (Exception e)
{
Console.WriteLine("The file could not be read:");
Console.WriteLine(e.Message);
Console.WriteLine(e);
}
}
