Add details to description

app.py

@@ -127,7 +127,19 @@ def train_model(n_train, noise):
     return fig, out_str
 
 title = "Single estimator versus bagging: bias-variance decomposition ⚖️"
-description = "This example illustrates and compares the bias-variance decomposition of the expected mean squared error of a single estimator against a bagging ensemble. "
+description = """This example illustrates and compares the bias-variance decomposition of the \
+                expected mean squared error of a single estimator (Decision Tree Regressor) \
+                against a bagging ensemble of Tree Regressors. \
+                
+                The dataset used for this demo is a one-dimensional synthetic dataset generated \
+                for a regression problem. In the top two figures, the blue line represents the true \
+                function and the blue dots represent the training data that are obtained by adding some \
+                random noise (user selected). The prediction of the models is represented by the red line. \
+                The average prediction of each estimator is presented in cyan.
+
+                In the two lower figures, we can see the decomposition of the expected mean squared error \
+                (red) into the bias (blue) and variance (green), as well as the noise part of the error (cyan).
+            """
 with gr.Blocks() as demo:
     gr.Markdown(f"## {title}")
     gr.Markdown(description)