private int GetRank(PerformanceIndicatorEntity experiment)
{
var positiveWeights = WeightConfiguration[0];
var negativeWeights = WeightConfiguration[1];
var rank = 0;
rank += experiment.Accuracy > 0 ? positiveWeights.Accuracy : experiment.Accuracy < 0 ? negativeWeights.Accuracy : 0;
rank += experiment.AccuracyBaseline > 0 ? positiveWeights.AccuracyBaseline : experiment.AccuracyBaseline < 0 ? negativeWeights.AccuracyBaseline : 0;
rank += experiment.AUC > 0 ? positiveWeights.AUC : experiment.AUC < 0 ? negativeWeights.AUC : 0;
rank += experiment.AUCPrecisionRecall > 0 ? positiveWeights.AUCPrecisionRecall : experiment.AUCPrecisionRecall < 0 ? negativeWeights.AUCPrecisionRecall : 0;
rank += experiment.AverageLoss > 0 ? positiveWeights.AverageLoss : experiment.AverageLoss < 0 ? negativeWeights.AverageLoss : 0;
rank += experiment.LabelMean > 0 ? positiveWeights.LabelMean : experiment.LabelMean < 0 ? negativeWeights.LabelMean : 0;
rank += experiment.Loss > 0 ? positiveWeights.Loss : experiment.Loss < 0 ? negativeWeights.Loss : 0;
rank += experiment.Precision > 0 ? positiveWeights.Precision : experiment.Precision < 0 ? negativeWeights.Precision : 0;
rank += experiment.PredictionMean > 0 ? positiveWeights.PredictionMean : experiment.PredictionMean < 0 ? negativeWeights.PredictionMean : 0;
rank += experiment.Recall > 0 ? positiveWeights.Recall : experiment.Recall < 0 ? negativeWeights.Recall : 0;
rank += experiment.TrainTime > 0 ? positiveWeights.TrainTime : experiment.TrainTime < 0 ? negativeWeights.TrainTime : 0;
return(rank < 0 ? 0 : rank > 11 ? 11 : rank);
}
private void GetSavedExperimentResults()
{
ExperimentResults = new ObservableCollection <PerformanceIndicatorEntity>();
try
{
using (var reader = new StreamReader(CurrentDirectory.GetAssetsDirectoryFolder(@"CSVFiles\ExperimentResults.csv")))
{
do
{
var line = reader.ReadLine();
if (line != null)
{
var values = line.Split(',');
var experimentResult = new PerformanceIndicatorEntity()
{
PreprocessingTechnique = values[0],
Process = values[1],
Accuracy = int.Parse(values[2]),
AccuracyBaseline = int.Parse(values[3]),
AUC = int.Parse(values[4]),
AUCPrecisionRecall = int.Parse(values[5]),
AverageLoss = int.Parse(values[6]),
LabelMean = int.Parse(values[7]),
Loss = int.Parse(values[8]),
Precision = int.Parse(values[9]),
PredictionMean = int.Parse(values[10]),
Recall = int.Parse(values[11]),
TrainTime = int.Parse(values[12]),
};
ExperimentResults.Add(experimentResult);
}
}while (!reader.EndOfStream);
}
}
catch (Exception ex)
{
_exceptionLogDataAccess.LogException(ex.ToString());
}
}
