FCE develops an open-source .NET library in the predictive analytics context. The project started February 2018 and focusses on following topics:

  • Nonlinear Regressions / Classifications
  • Machine Learning
  • Clustering

Especially, a bundle of nonlinear regression techniques building the core of any forecasting algorithm and can be used in anomaly detecion, pattern recognition and time series forecast applications.

We share basic and state of the art machine learning algorithms with the .NET community and have the goal to establish our competence on that subject.

// Please choose the path of a data file (csv)
Tuple<List<double[]>, List<double[]>> tuple = FCECalculus.Access.ReadCSVFile(@"C:\banana.csv");
List<double[]> X = tuple.Item1;
List<double[]> Y = tuple.Item2;

// Please choose a RegressionType such as Closest
ARegression regression = FCECalculus.ENums.RegressionType.Closest.ToRegression();
// Training Procedure
regression.Training(X, Y);

// Fill x with data that you would like to evaluate...
double[] x = new double[X[0].Length];
double[] yHat = regression.Evaluate(x, X, Y);