<![CDATA[Determining a healthy lifestyle is an important issue in public health, especially in efforts to prevent chronic diseases. The classification process is carried out by constructing a decision tree that divides data into lifestyle classes (healthy/unhealthy) recursively, based on the features that provide the best separation. The results show that the CART model is able to identify significant lifestyle patterns with fairly high classification accuracy, as well as provide a clear understanding of user factors that contribute to healthy lifestyle status. This approach supports decision making in data-based health promotion programs. From a total of 70 data, the average target value is 0.146. The tree will divide the data based on whether the Feature value is ≤ 3.25. If true to the left, if false to the right. This node shows 49 samples from the root following the condition Feature <= 3.25 and is now divided again based on Feature <= 2.538. This process continues recursively until it reaches the leaf node. There is only 1 sample, with target value = 0.3, so there is no variance (squared_error = 0). If new data enters this branch, the model will predict 0.3 as the regression value.]]>
Copyrights © 2025
