15.   DecisionMakerPoly63 Class Library
15.1   DecisionMakerPoly63 Interface
15.2   Link to Class Library
15.3   Prediction Buttons
15.4   Test Run

15.   DecisionMakerPoly63 Class Library

This chapter introduces the class library, Attrasoft.PolyApplet63.DecisionMakerPoly63, which is the library for function inference using the PolyNet algorithm; the last chapter uses the Abm algorithm. In this chapter, we will build prediction software similar to Attrasoft DecisionMaker 2.5, but using the PolyNet algorithm. The prediction operation includes two steps:

The only difference between this chapter and the last chapter is to change a few commands.

Figure 15.1   The PolyDecisionMaker Form.

15.1   DecisionMakerPoly63 Interface

The DecisionMakerPoly63 class library implements the following interface:

 public interface I_DecisionMakerPoly63
  bool openDecisionMakerTrainFile ();
  bool openDecisionMakerRecoFile ();
  bool openOutputFile ();

  bool decisionMakerPolyLinearInt ();
  bool decisionMakerPolyLinearReal  ();

In addition, the class library has the following properties:
 I/O Files:
  public string DecisionMakerTrainName
  public string DecisionMakerRecoName
  public string OutputName
 Presentation Layer parameters:
  public double Emptyfield
 Neural Layer parameters:
  public int Sensitivity
  public int Blurring
  public int InternalWeightCut
15.2   Link to Class Library

The class library is:

The class in this library will be:
To include the class library in the project, To use the class library, add:
using Attrasoft.PolyApplet63.DecisionMakerPoly63;
To declare an object, add:
public Attrasoft.PolyApplet63.DecisionMakerPoly63.DecisionMakerPoly63 z
= new Attrasoft.PolyApplet63.DecisionMakerPoly63.DecisionMakerPoly63 (richTextBox1);
Now DecisionMakerPoly63 object, z, is ready to use.

15.3   Prediction Buttons

There are 2 commands in Figure 15.1; each button will simply use one of the 2 commands:

 private void button8_Click(object sender, System.EventArgs e)
   z.decisionMakerPolyLinearInt ();

 private void button9_Click(object sender, System.EventArgs e)
   z.decisionMakerPolyLinearReal ();

15.4   Test Run

Step 1. Files.

The following are done automatically:

DecisionMaker Train:                       cancer1a.txt:
DecisionMaker Recognition:             cancer1b.txt:
Train:                                              example1a.txt
Recognition:                                     example1c.txt
Neural Output:                                 example1d.txt
Output:                                            example2c.txt

Step 2. Click the �Int Linear� button to see the results.

Step 3. Click the �Real Linear� button to see the output in Real Numbers instead of Integers.

We should see the same results as in chapter 11.