Research Article
An Efficient Normalized Rank Based SVM for Room Level Indoor WiFi Localization with Diverse Devices
Table 2
Algorithms comparison based on the percentage of the well-recognized positions; the radio map defined upon RSSI values.
| Device type | Classification type | SVM | Artificial Neural Network | Naïve Bayes | Random Forest | KNN () | L2-R L2-N |
| Oppo A31c | | 97.50% | | | | BBK | 94.17% | | | | | Coolpad 8730L | 90.62% | 97.71% | | | | Gionee | 99.17% | | | | | HTC One E8 | | 91.88% | | | | Huawei GRA-CL00 | 97.92% | 93.33% | | | | Lenovo A788t | 94.58% | | | | | Meizu | 99.58% | 99.79% | 95.00% | | 99.58% | Oppo R7c | 98.12% | | | | | Xiaomi | | | | | | Xiaomi Cancro | | | | | | Samsung klteduoszn | 92.92% | 98.75% | | | | Meizu M2 note | 97.71% | 97.71% | | | | Samsung trlteduosctc | 93.95% | 97.29% | | | | BBK Vivo | 96.87% | | | | |
| Standard device accuracy | 91.87% | 97.50% | | | | Recall (ratio) | 0.92 | 0.97 | 0.77 | 0.85 | 0.87 | Precision (ratio) | 0.95 | 0.97 | 0.85 | 0.93 | 0.93 |
|
|
We put in bold the percentages lower than 90% for easiness of observation.
|