Datasets
Standard Dataset
2000_MTJ_samples
- Citation Author(s):
-
ShailendraYadav
Indian Institute of Technology Roorkee - Submitted by:
- Shailendra Yadav
- Last updated:
- Tue, 09/24/2024 - 13:32
- DOI:
- 10.21227/q2s5-y479
- Data Format:
- License:
32 Views- Categories:
- Keywords:
0 ratings - Please login to submit your rating.
Abstract
Two thousand MTJ samples in total were included in the dataset for analysis; First, shuffle the dataset randomly to eliminate bias, then split it into ten equal folds of 200 samples each. For each iteration of cross-validation, use nine folds for training and one for testing, rotating the test fold across all ten groups so that every sample is tested once.
Instructions:
Step-1: Genereate the 2000 STT-MTJ data from SPICE simulation.
Step-2: Four input variation (Alpha, TMR, Tunnel barrier thickness, Free layer thickness) and Switching/critical current required to switch the state of STT-MTJ.
step-3: Shuffle the dataset randomly to eliminate bias,
then split it into ten equal folds of 200 samples each.
For each iteration of cross-validation, use nine folds for training and one for testing,
rotating the test fold across all ten groups so that every sample is tested once.
Documentation
| Attachment | Size |
|---|---|
 readme.txt | 574 bytes |
Comments
no