Data Science release notes encapsulate updates to tools contained with the Campus Analytics product.
Early Warning is an opt-in framework that utilizes existing, anonymized Campus data to transform raw data into knowledge. Using this system, Campus can provide this knowledge and insight to customers as meaningful predictive analysis for an individual student or students.
No personally identifiable information is made available to the Data Science system, rather the Data Science system is used to train algorithms using anonymous data, after which these algorithms are applied to the district’s data where the results are connected to specific students. That information is available only to district and state users with rights.
Below is a list of recent Data Science releases:
Early Warning Modeling - 0.30.6 - December 2021
New Model Available (DASC-5899)
Infinite Campus has developed and deployed a new Early Warning national prediction model. This model incorporates student outcomes for the 2020-2021 academic year.
Existing Early Warning customers may see a change to existing GRAD and category scores as these scores are updated to utilize the improved prediction model. This update requires no action by the user; however, users are encouraged to review Early Warning data to see how updated scores have impacted students in their school.
Early Warning Modeling - 0.27.0 - March 2021
New Model Available (DASC-4776)
Infinite Campus has developed and deployed a new Early Warning national prediction model. This model incorporates student outcomes for the 2019-2020 academic year in addition to some process changes regarding how districts are evaluated for Early Warning eligibility. This change in process is a net gain in the number of districts that are eligible for Early Warning. This change in process does not impact current Early Warning customers.
Existing Early Warning customers may see a change to existing GRAD and category scores as these scores are updated to utilize the improved prediction model. This update requires no action by the user; however, users are encouraged to review Early Warning data to see how updated scores have impacted students in their school.
Early Warning Modeling - 0.25.0 - January 2020
New Model Available (DASC-3434)
Infinite Campus has developed and deployed a new Early Warning national prediction model. This model incorporates student outcomes for the 2018-2019 academic year in addition to some process changes regarding how districts are evaluated for Early Warning eligibility. This change in process is a net gain in the number of districts that are eligible for Early Warning. This change in process does not impact current Early Warning customers.
Existing Early Warning customers may see a change to existing GRAD and category scores as these scores are updated to utilize the improved prediction model. This update requires no action by the user; however, users are encouraged to review Early Warning data to see how updated scores have impacted students in their school.
Early Warning Modeling - 0.22.0 - June 2019
National Model Now Available (DASC-2661)
Infinite Campus has developed and deployed a new Early Warning national prediction model. This model incorporates significantly more students and data across all of our available customers, leading to a greatly improved predictive quality of Early Warning GRAD scores and category scores.
Kentucky customers will see a change to existing GRAD and category scores as these scores are updated to utilize the improved prediction model. This update requires no action by the user; however, users are highly encouraged to review Early Warning data to see how updated scores have impacted students in their school.
Early Warning Messaging - 0.20.0 - May 2019
Cache Scheme Decoupled from SIS Release Process (DASC-2409)
The Early Warning cache schema (cache data/calculation used to build and release new Early Warning models) has been decoupled from the SIS release process. This decoupling gives Campus control of cache releases by the central EW messaging application and allows all District Edition cache versions to stay up-to-date. Benefits of this decoupling include:
- The ability to support a range (min/max) of Campus versions
- Ability to deploy asynchronously with the SIS Campus release
- No Tomcat bounce is required
- Support for Cloud and Cloud Choice hosting models
- Deployments can be made based on internal decisions and not via a customer request
This update does NOT impact customers or existing GRAD Scores.
Early Warning API - 0.3.0 - October 2018
Updates Made to Stabilize Beginning of Year GRAD Scores (DASC-1612)
Enhancements were made to Campus Early Warning to improve how the system handles attendance and grade percent calculations when a student has little data in the current year. Previously, students who experienced wide variations in Gradebook or attendance records at the beginning of the year also saw large GRAD Score changes. This update helps to stabilize GRAD Scores.
This update may change existing GRAD Score values.
Null Value Modification (DASC-1668)
Modifications were made for how the system handles null values.
This update may change existing GRAD Score values.
Early Warning Enhancements for Smaller Districts (DASC-1699)
Performance improvements were made which impact small districts.
These corrections improve the accuracy of GRAD Scores for small districts and may have an impact on existing scores. In particular, some students’ GRAD Scores could be significantly lower (though more accurate) after this update.
Early Warning Model - 0.1.101 - August 2018
Early Warning Algorithm Update Published (Closed Beta) (DASC-1445)
An Early Warning algorithm update has been published for closed beta customers. This update corrects minor bugs and includes improvements to the accuracy of predicted GRAD scores.
