Use Case 3: Using ML To Transform Learner Outcomes

A brief description of what this project does and who it's for

Description of Files

1. joinedData_assessmentStudentAssessor3_clean Anonymized.zip

• Raw data for the ML model
• Columns for the data
  • Anonymized data - StudentId, AssessmentId, OrganizationId, AssessorId
  • Assessment data - A1 through Z28. Different skill areas of Basic Learner Skills, specified in ABLLS-R Guide 2018 pg 30-49
  • Assessment metadata - FirstAssessment_byStudent, Col UC through UV

2. ABLLS-R Guide 2018.zip

• Provide details about the scoring model
• Case studies to provide examples to show scoring for sampel assessments
• The purpose of this model is to
  • identify critical skills that are in need of intervention
  • provide a method for identifying a child's specific skill
  • provide a curriculum guide for an educational program for a child

3. Normative_Report_Hackathon23.pdf

• This was a study done with 53 students whose data is provided in the two files below
• examplesNeuroTypical_ABLLSR.csv
  • Similar data as in the anonymized data set, but from teh study above
  • These individual assessment scores could help:
    • offer a training data set
    • characterize skill acquisition pathways
    • give timelines in skill acquisition pathways
• normativeBenchmarkAsessments.csv
  • This provides average assessment for each category for ages 2 years, 3 years, 4 years, and 5 years
  • There is not enough data for normalized values for 1 year patients
  • Assessment score of 0-4 is normalized from 0-1 (dividing by 4)

Sample Problem Statements

Some sample problem statements are provided below that you can adopt as the objective of your ML model. However, you are free to create a new problem statement that you believe the data in joinedData_assessmentStudentAssessor3_clean Anonymized.csv can solve

1. Problem Statement #1 - Identify peer group for a patient to compare progress against and identify the most important features in the dataset
2. Problem Statement #2 - Determine whether progress is being made based on assessment
3. Problem Statement #3 - Identify if the patient is in the right program Hard as dataset does not have program information.

Live Web Page

https://jolly-desert-01b5c5610.3.azurestaticapps.net/

Team Summary

• Dhruv BHatnagar, Presentation
• Rishi Bhatnagar, Presentation
• Noha Elprince, Machine Learning Expert
• Olive Kingsley, Data Scientist
• Anita Mohanani, QA learning AI
• Ryan Carpenter, Jack of all trades!
• Adam Pryce, DevOps, and Web Stack
• Halaa Menasy, Tech Lead
• Peter Shand, Tech Lead
• Hanen Bondka, Shyam Sundar, Parvesh Mamgain - Data Munging

Zoom room

Join Zoom Meeting https://morganstanley.zoom.us/j/5293623419

One tap mobile +13462487799,,5293623419# US (Houston) +16694449171,,5293623419# US