4 Preprocess Spatial Data

You can prepare, process, and augment source data for spatial analysis using Oracle Spatial AI.

You can use the processing methods provided by Spatial AI for filling missing values, scaling data, engineering new features, and splitting the source dataset into training, validation, and test datasets.

Note that data preprocessing may vary for different scenarios. Also, some methods work only for numerical data, and the others work only for categorical variables. However, the machine learning algorithms work with numbers, and the preprocessing methods aim to provide "good" data to the algorithms to produce meaningful results.

The following preprocessing methods are implemented in Oracle Spatial AI.

Topics:

• Spatial Lag Transformer
• Categorical Lag Transformer
• Spatial Coordinates Transformer
• Spatial Imputer
• Splitting Datasets