Examples.md

-# Geological data example
-
-## Data Description
+# Example methodology description
 
-The data set contains 9 variables with 442 samples. The target variable for prediction in the following example is 'Depth'. The variable is also used to visually evaluate sampling quality via distribution plot.
-
-This and the next example consider the following combinations of hyperparameters used for Bayesian network learning:
+Both examples consider the following combinations of hyperparameters used for Bayesian network learning:
 
 * K2 metric;
 * K2 metric with gaussian mixtures (GMM);
@@ -13,25 +9,63 @@ This and the next example consider the following combinations of hyperparameters
 
 All the examples are executed using cross-validation.
 
-## K2 metric sampling example
+# Geological data example
+
+## Data Description
+
+The data set contains 9 variables with 442 samples. The target variable for prediction in the following example is 'Depth'. The variable is also used to visually evaluate sampling quality via distribution plot.
+
+## Sampling
+
+### K2 metric sampling example
 
 ![k2](https://user-images.githubusercontent.com/86363785/188129119-dfa62b6d-b1fd-4e63-aa75-fb7aafba95a1.png)
 
-## Sampling with K2 + GMM example
+### Sampling with K2 + GMM example
 
 ![geo_k2_gmm](https://user-images.githubusercontent.com/86363785/188129748-ce239eb4-bbab-43f0-9d80-c92483f27613.png)
 
-## Sampling with K2 + GMM + logit nodes example
+### Sampling with K2 + GMM + logit nodes example
 
 ![geo_k2_gmm_logit](https://user-images.githubusercontent.com/86363785/188129774-a3695199-776d-493f-8a9c-bf78125f03fb.png)
 
-## K2 with initial structure sampling
+### K2 with initial structure sampling
 
 ![geo_k2_expert](https://user-images.githubusercontent.com/86363785/188129863-b8777153-eb31-4e8f-b8bf-b87e7c959035.png)
 
 
 # Social data example
 
-The second example is similar to the previous one, but carried out on different data set. Social data set consists of 30000 anonymous bank records with 9 variables each, bayesian networks were learnt on a sample with 2000 records. 
+## Data Description
+
+The second example is similar to the previous one, but carried out on different data set. Social data set consists of 30000 anonymous bank records with 9 variables each, bayesian networks were learnt on a sample with 2000 records. The target variable is 'mean_tr' which is mean transaction of client. 
+
+## Sampling 
+
+### K2 metric sampling example
+
+![socio_k2](https://user-images.githubusercontent.com/86363785/188132481-2ae015e4-69a0-4025-84ef-c96aad6dd98e.png)
+
+
+### Sampling with K2 + GMM example
+
+![social_k2_gmm](https://user-images.githubusercontent.com/86363785/188132496-e49ebf7d-d603-406a-a199-8cbc3e3c256c.png)
+
+
+### Sampling with K2 + GMM + logit nodes example
+
+![social_k2_gmm_logit](https://user-images.githubusercontent.com/86363785/188132505-72b257f8-fb38-47b3-ad19-a670bd31c6d9.png)
+
+
+### K2 with initial structure sampling
+
+![socio_expert](https://user-images.githubusercontent.com/86363785/188132518-5d463979-3b9d-46b0-8cda-5cef791ec98c.png)
 
+# Prediction MSE table for both examples 
 
+| Hyperparameters combinations | Geological data MSE | Social data MSE |
+|------------------------------|---------------------|-----------------|
+| K2                           | 1014.59             | 6066.5          |
+| K2 + GMM                     | 974.35              | 5149.5          |
+| K2 + GMM + logit             | 1018.84             | 6657.93         |
+| K2 + initial structure       | 1056.06             | 12506.47        |
\ No newline at end of file