Which node applies transformations to dataset variables for better distribution?

The AutoNeural Node is the correct choice for applying transformations to dataset variables to achieve better distribution. This node is designed to automatically preprocess the input variables by applying various transformations that can help improve the model's performance. This can include scaling, normalization, and handling of outliers, among other techniques. These transformations aim to ensure that the input features are on a similar scale or have properties that enhance learning, which is particularly important for neural networks.

In contrast, the other nodes listed focus on different aspects of modeling. The Gradient Boosting Node is primarily concerned with creating an ensemble model through boosting, which combines weak learners but does not inherently focus on transforming the input features for distribution improvement. The Decision Tree Node creates decision rules based on features, while the DMNeural Node is specific to neural networks but does not inherently apply transformations like feature scaling or normalization unless specifically configured to do so.

Thus, the AutoNeural Node stands out as the specific tool designed explicitly for improving the distribution of dataset variables as part of its preprocessing capabilities.