Spinner logo OtasAI


Did You Know?

Handling missing values for machine learning


Handling missing values is a crucial step in data preparation, as it ensures the integrity and completeness of the dataset, which is essential for accurate analysis and model training. OtasML, a visual machine learning tool, offers a robust Handle Missing Values feature within its data preparation model. This tool provides various methods to handle missing data, allowing users to choose the most appropriate strategy for their specific needs. This article explores the different methods available and how they can be configured to optimize your machine learning workflows.


The Handle Missing Values tool in OtasML offers multiple strategies for dealing with missing data, enabling users to maintain the quality of their datasets. Below are the key configurations and options available:


  • Default Value: None
  • Description: This feature allows users to populate empty rows with values in their dataset using various methods. The available methods include:
    • Backfill: Fills the missing values in the selected subsets by using the next non-missing value in a backward direction.
    • Bfill: Replaces the NULL values with the values from the next row, effectively backfilling the missing data.
    • Pad: Fills or replaces missing values in selected subsets with the most recent non-missing value along a specified axis. This method is useful for forward-filling missing values, carrying forward the last known value.
    • Ffill: Similar to Pad, it fills missing values in selected subsets using the most recent non-missing value in a forward direction.
    • Drop: Removes rows containing missing (NaN) values from the selected subsets, useful for eliminating incomplete records.
    • Fill0: Replaces rows containing missing (NaN) values with zero values in the selected subsets.
    • Median: Replaces rows containing missing (NaN) values with the median values of the selected subsets.
    • Mean: Replaces rows containing missing (NaN) values with the mean values of the selected subsets.


  • Default Value: None
  • Description: The Subset option allows users to select specific columns for identifying and filling missing values. This ensures that only the desired columns are processed, providing more control over the data preparation step.


  • Default Value: None
  • Description: This parameter controls how many consecutive missing values are filled before the filling process stops. It allows users to limit the number of NaN values that are filled, ensuring that the filling process does not extend beyond a reasonable range. Note that in the Drop method, the Limit specifies the minimum number of non-NA values that must be present in a row for it to be kept.

Interactive Buttons: Preview and Save

To enhance user experience and provide greater control over the handling of missing values, the tool includes two essential buttons:

  • Preview: This button allows users to see the effects of the selected method for handling missing values in real-time without permanently applying the changes. By clicking Preview, users can visually assess how the dataset will be altered based on the current configurations, ensuring that the chosen method is appropriate before committing to any changes.
  • Save: Once users are satisfied with their configurations and the preview results, they can click the Save button to permanently apply their chosen settings. This action saves the configuration, which will then be applied to the data during the training process, ensuring that the handling of missing values aligns with the user's expectations and requirements.


The Handle Missing Values tool in OtasML provides a comprehensive solution for dealing with missing data, ensuring the integrity and completeness of datasets. By offering a variety of methods to handle missing values and the ability to selectively apply them to specific columns, users can effectively tailor the data preparation step to their specific needs. The inclusion of interactive Preview and Save buttons further enhances control and confidence in the missing value handling process. OtasML continues to empower users with intuitive and powerful tools, making data preparation a seamless and integral part of the machine learning workflow.


A+ A-