Welcome to the 2020.02.26 product release.  This article provides an overview of the product innovations and improvements delivered on 26 February 2020.  For all future release plans check out our release calendar for 2020.  The release notes are structured as follows:

User Experience

OneAi Innovations

 Bugs, Performance & Platform Improvements

  • Data Ingestion Bugs Fixed

  • General Bugs Fixed and Minor Improvements

User Experience

Update on Beta Release of New User Experience (UX) & Storyboards

  • We are rounding the final turn and getting close to closing the beta period for the new UX.  In the latest release we smashed lots of bugs and added a bunch of new features to deliver on most of the key product features.  We already have a number of customers that have fully migrated to the new UX and rolled out to their users.  Please continue to send through your feedback.

  • To help you deploy the new UX at scale we added a new option to enable the new UI for your entire company and also an option to enable the new UX just for specific roles. (ref 4003 & 4004)

  • To enable the new UX for the entire company there is a new option in the Company Admin page:

  • To enable the new UX for specific roles  there is a new option in the Application Access Roles permissions page called "Use2020UserExperience":

OneAi Innovations

  • We’ve released the first version or PCI ordinal encodings which allow you to express ordered categorical columns for better EDA and potentially model performance.

  • We’ve expanded some EDA report functionality to give additional visuals to help you understand variables.

   Violin Plots

Scatter Plots

In this release we have delivered a bunch of improved content and visualizations for the Results Summary model performance and explanation graphs. 

New Regression Plots

  • Residuals: This plot shows the difference between the predicted value and the actual value.

  • Prediction Error: This plot shows the predicted value plotted against the actual value.

  • Prediction Distribution Plot: This is a distribution plot of the predicted values with a normal curve overlay to give context for outliers.

  • Outlier Table: This table shows individual outlier values.

Classification

  • Class Balance:  This graph shows the number of instances for each class label.

  • Class Prediction Error: This graph shows the breakdown of the label predictions vs. the actual value between the train and test set.

  • Confusion Matrix - OneAi already includes a comprehensive Classification Report, but due to popular demand we have also added the Confusion Matrix to describe the performance of a classification model.  

Bugs, Performance & Platform Improvements

Data Ingestion Bugs Fixed

  • To add greater control over your ability to handle very large datasets, or just reload a shorter window of data, we have added the ability to load SuccessFactors API data for a defined time period up to the current day using the One Model connector. (ref 3692)

  • Improved the resilience of the data loading process when the environment is under heavy load. (ref 4009 & 4014)

General Bugs Fixed and Minor Improvements

  • Fixed an error where Data Destinations from Processing Scripts could not be setup if a column was configured as Double or Single in the Data Model.(ref 3998)

  • Added support for GREATEST() and LEAST() functions in Processing Script. (ref 3937)

  • We fixed a minor UI misalignment in Dashboard Permissioning within the Data Access Roles permission page (ref 3611)

  • We tightened up the control of a user’s ability to publish Dashboards/Storyboards to ensure users can only publish to roles that they have access to.  (ref 4008)

Did this answer your question?