GradeCast AI-powered Assessment Insight Platform (Artificial Neural Network Prediction + Rasch Diagnostics)

1. Upload & Column Mapping

Item Descriptive CSV

Include: ItemID, Construct, Format ... (any columns you want).

Student Responses CSV

Long: (StudentID, ItemID, Score). Wide: row=student, columns=items.

Item file: - | Score file: -

2. Feature Selection (for ANN)

Choose which columns from the Item CSV participate in the model, and set encoding type.

3. Model Structure & Validation

Model Structure

Hidden Layers

Validation Configs (add more)

4. Start Training

5. Training Results

6. Predict New Items

7. IRT Calibration & Wright Map