Changes between Version 20 and Version 21 of UsersSamples

Timestamp:

02/18/11 13:19:57 (14 years ago)

Author:

gkronber

Comment:

--

UsersSamples

@@ -251 +251 @@
 === Genetic programming - Symbolic Regression (Tower)===
 [attachment:SGP_SymbReg.hl]
-Example for a simple genetic programming algorithm to create a regression model for the estimation of a product quality parameter in an industrial chemical process. The original dataset was downloaded from http://vanillamodeling.com/realproblems.html. 
+
+Example for a simple genetic programming algorithm to create a regression model for the estimation of a product quality parameter in an industrial chemical process. The original dataset was downloaded from http://vanillamodeling.com/realproblems.html.
+
+ * Population size: 500
+ * Generations: 100
+ * Tournament selection: group size=7
+ * Mutation rate: 15%
+ * Function set: +, *, -, /, average, log, power, root, exponential
+ * Terminal set: constants, x01, x02, x03, x04, x05, x06, x08, x09, x10, x12, x14, x15, x17, x18, x19, x20, x22, x23, x24
+ * Fitness function: R² of predicted and original tower response
 
 ----
@@ -257 +266 @@
 [=#GPSymbolicRegressionMackeyGlass]
 === Genetic programming - Symbolic Regression (Mackey-Glass)===
+[attachement:SGP_SymbReg-Mackey-Glass.hl]
+
+Example for a simple genetic programming algorithm to create a dynamic model for the one-step prediction of the chaotic Mackey Glass (T=17) time series. The original dataset was downloaded from http://www.bme.ogi.edu/~ericwan/data.html.
+
+ * Population size: 500
+ * Generations: 100
+ * Tournament selection: group size=7
+ * Mutation rate: 15%
+ * Function set: +, *, -, /, average, log, power, root, exponential
+ * Terminal set: constants, x(t-90) .. x(t-1)
+ * Fitness function: R²(xPred(t), xOrig(t))
+
+
 ----
 
@@ -262 +284 @@
 === Genetic programming - Symbolic Classification (Wisconsin)===
 [attachment:SGP_Classification-WDPC.hl]
+
 Example for a simple genetic programming algorithm to create a classification model for the estimation of malignant or benign tumor diagnosis based on features extracted through analysis of tumorous cells in a tissue sample. The original dataset was downloaded from the UCI Machine Learning Repository (http://archive.ics.uci.edu/ml/datasets/Breast+Cancer+Wisconsin+(Diagnostic)). The algorithm uses all available GP manipulation operators, single-point cross-over and squared Pearson's correlation coefficient evaluation.
 
@@ -271 +294 @@
 === Genetic programming - Symbolic Classification (Mammography)===
 [attachment:SGP_SymbClass-Mammographic.hl]
+
  A genetic programming algorithm to create a classification model for the prediction of malignant or benign tumor diagnosis based on features extracted through a non-invasive mammography breast cancer screening. Original dataset stems from the UCI Machine Learning Repository (http://archive.ics.uci.edu/ml/datasets/Mammographic+Mass).
 
@@ -278 +302 @@
 === Genetic Programming - Even Parity (4 inputs) ===
 [attachment:SGP_Boolean-Even-Parity-4.hl]
+
  A genetic programming algorithm to synthesize the boolean even parity function with 4 binary inputs
 
@@ -297 +322 @@
 === Genetic Programming - Multiplexer (11 inputs ===
 [attachment:SGP_Boolean-Multiplexer-11.hl]
+
  A genetic programming algorithm to synthesize the boolean multiplexer function with 11 binary inputs