Image Classification on MNIST (test) (Accuracy and Hardware Metrics)
95.1AccuracyDTM
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| DTMPlatform=XC7Z020, Algorithm=TM, Purpose=Training2026.03 | 95.1 | 43,497 | 33,256 | 138 | 6 | 95.1 | 4.36 | |
| BTHOWeNPlatform=XC7Z020, Algorithm=WiSARD, Purpose=Inference2026.03 | 93.4 | 15,756 | 3,522 | 0 | 0 | 250 | 0.3 | |
| TsetlinWiSARD-300Platform=XC7Z020, Algorithm=WiSARD, Purpose=Training2026.03 | 92.1 | — | — | — | — | — | — | |
| LogicWiSARDPlatform=XC7Z010, Algorithm=WiSARD, Purpose=Inference2026.03 | 91 | 2,142 | 163 | 0 | 0 | 560 | 0.12 | |
| Hashed WNNPlatform=XC7Z020, Algorithm=WiSARD, Purpose=Training2026.03 | 90.73 | 9,286 | 4,568 | 129 | 5 | 8.3 | 0.53 | |
| TsetlinWiSARD-150Platform=XC7Z020, Algorithm=WiSARD, Purpose=Training2026.03 | 88.02 | 33,596 | 25,927 | 2 | 1 | 6.4 | 1.56 |