Middle-exclusion effect on a neural network trained to classify lncRNA. The network architecture was an MLP. The features were k-mer profiles of RNA sequences. The data, features and training regimes were as in previous tables. The middle exclusion filter was applied to (A) all the data, (B) none of the data, (C) the training subsets only or (D) the validation subsets only
| . | A . | B . | C . | D . |
|---|---|---|---|---|
| Train filter: validation filter: . | Middle excl. middle excl. . | None none . | Middle excl. none . | None middle excl. . |
| Accuracy | 68.7 ± 2 | 62.2 ± 3 | 61.7 ± 1 | 70.9 ± 2 |
| F1 | 63.0 ± 3 | 60.1 ± 4 | 61.3 ± 1 | 70.7 ± 2 |
| MCC | 0.390 | 0.254 | 0.236 | 0.419 |
| AUPRC | 78.7 ± 1 | 68.0 ± 3 | 66.4 ± 2 | 78.4 ± 2 |
| AUROC | 77.2 ± 1 | 67.8 ± 3 | 66.1 ± 1 | 77.6 ± 2 |
| . | A . | B . | C . | D . |
|---|---|---|---|---|
| Train filter: validation filter: . | Middle excl. middle excl. . | None none . | Middle excl. none . | None middle excl. . |
| Accuracy | 68.7 ± 2 | 62.2 ± 3 | 61.7 ± 1 | 70.9 ± 2 |
| F1 | 63.0 ± 3 | 60.1 ± 4 | 61.3 ± 1 | 70.7 ± 2 |
| MCC | 0.390 | 0.254 | 0.236 | 0.419 |
| AUPRC | 78.7 ± 1 | 68.0 ± 3 | 66.4 ± 2 | 78.4 ± 2 |
| AUROC | 77.2 ± 1 | 67.8 ± 3 | 66.1 ± 1 | 77.6 ± 2 |
Middle-exclusion effect on a neural network trained to classify lncRNA. The network architecture was an MLP. The features were k-mer profiles of RNA sequences. The data, features and training regimes were as in previous tables. The middle exclusion filter was applied to (A) all the data, (B) none of the data, (C) the training subsets only or (D) the validation subsets only
| . | A . | B . | C . | D . |
|---|---|---|---|---|
| Train filter: validation filter: . | Middle excl. middle excl. . | None none . | Middle excl. none . | None middle excl. . |
| Accuracy | 68.7 ± 2 | 62.2 ± 3 | 61.7 ± 1 | 70.9 ± 2 |
| F1 | 63.0 ± 3 | 60.1 ± 4 | 61.3 ± 1 | 70.7 ± 2 |
| MCC | 0.390 | 0.254 | 0.236 | 0.419 |
| AUPRC | 78.7 ± 1 | 68.0 ± 3 | 66.4 ± 2 | 78.4 ± 2 |
| AUROC | 77.2 ± 1 | 67.8 ± 3 | 66.1 ± 1 | 77.6 ± 2 |
| . | A . | B . | C . | D . |
|---|---|---|---|---|
| Train filter: validation filter: . | Middle excl. middle excl. . | None none . | Middle excl. none . | None middle excl. . |
| Accuracy | 68.7 ± 2 | 62.2 ± 3 | 61.7 ± 1 | 70.9 ± 2 |
| F1 | 63.0 ± 3 | 60.1 ± 4 | 61.3 ± 1 | 70.7 ± 2 |
| MCC | 0.390 | 0.254 | 0.236 | 0.419 |
| AUPRC | 78.7 ± 1 | 68.0 ± 3 | 66.4 ± 2 | 78.4 ± 2 |
| AUROC | 77.2 ± 1 | 67.8 ± 3 | 66.1 ± 1 | 77.6 ± 2 |
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
