| Notation . | Description . |
|---|---|
| |$\mathcal{G}=\left (\mathcal{V}, \mathcal{E}, w\right )$| | A graph |$\mathcal{G}$| constructed of nodes |$\mathcal{V}$| and edges |$\mathcal{E}$|, with |$w$| denoting the set of edge weights |
| |$v_{i}$| | The |$i$|-th node |
| |$n$| | The number of nodes |
| |$\rho $| | A metapath |
| |$R$| | Edge types |
| |$\mathcal{A}$| | Node types |
| |$s\left (v_{i},v_{j}|\rho \right )$| | The proximity between node |$v_{i}$| and node |$v_{j}$| in path instance |$\rho $| |
| |$\mathcal{S}=\left \lbrace s_{i}\right \rbrace _{n}^{i=1}$| | The herb-protein transfer probability matrix and the input data for the deep autoencoder |
| |$K$| | The number of layers of the deep autoencoder |
| |$\mathcal{W}^{k}, \hat{\mathcal{W}}^{k}$| | The |$k$|-th layer weight matrix of the deep autoencoder |
| |$b^{k}, \hat{b}^{k}$| | The |$k$|-th layer biases |
| |$\boldsymbol{y}_{i}^{k}$| | The |$k$|-th layer hidden representation for |$v_{i}$| |
| Notation . | Description . |
|---|---|
| |$\mathcal{G}=\left (\mathcal{V}, \mathcal{E}, w\right )$| | A graph |$\mathcal{G}$| constructed of nodes |$\mathcal{V}$| and edges |$\mathcal{E}$|, with |$w$| denoting the set of edge weights |
| |$v_{i}$| | The |$i$|-th node |
| |$n$| | The number of nodes |
| |$\rho $| | A metapath |
| |$R$| | Edge types |
| |$\mathcal{A}$| | Node types |
| |$s\left (v_{i},v_{j}|\rho \right )$| | The proximity between node |$v_{i}$| and node |$v_{j}$| in path instance |$\rho $| |
| |$\mathcal{S}=\left \lbrace s_{i}\right \rbrace _{n}^{i=1}$| | The herb-protein transfer probability matrix and the input data for the deep autoencoder |
| |$K$| | The number of layers of the deep autoencoder |
| |$\mathcal{W}^{k}, \hat{\mathcal{W}}^{k}$| | The |$k$|-th layer weight matrix of the deep autoencoder |
| |$b^{k}, \hat{b}^{k}$| | The |$k$|-th layer biases |
| |$\boldsymbol{y}_{i}^{k}$| | The |$k$|-th layer hidden representation for |$v_{i}$| |
| Notation . | Description . |
|---|---|
| |$\mathcal{G}=\left (\mathcal{V}, \mathcal{E}, w\right )$| | A graph |$\mathcal{G}$| constructed of nodes |$\mathcal{V}$| and edges |$\mathcal{E}$|, with |$w$| denoting the set of edge weights |
| |$v_{i}$| | The |$i$|-th node |
| |$n$| | The number of nodes |
| |$\rho $| | A metapath |
| |$R$| | Edge types |
| |$\mathcal{A}$| | Node types |
| |$s\left (v_{i},v_{j}|\rho \right )$| | The proximity between node |$v_{i}$| and node |$v_{j}$| in path instance |$\rho $| |
| |$\mathcal{S}=\left \lbrace s_{i}\right \rbrace _{n}^{i=1}$| | The herb-protein transfer probability matrix and the input data for the deep autoencoder |
| |$K$| | The number of layers of the deep autoencoder |
| |$\mathcal{W}^{k}, \hat{\mathcal{W}}^{k}$| | The |$k$|-th layer weight matrix of the deep autoencoder |
| |$b^{k}, \hat{b}^{k}$| | The |$k$|-th layer biases |
| |$\boldsymbol{y}_{i}^{k}$| | The |$k$|-th layer hidden representation for |$v_{i}$| |
| Notation . | Description . |
|---|---|
| |$\mathcal{G}=\left (\mathcal{V}, \mathcal{E}, w\right )$| | A graph |$\mathcal{G}$| constructed of nodes |$\mathcal{V}$| and edges |$\mathcal{E}$|, with |$w$| denoting the set of edge weights |
| |$v_{i}$| | The |$i$|-th node |
| |$n$| | The number of nodes |
| |$\rho $| | A metapath |
| |$R$| | Edge types |
| |$\mathcal{A}$| | Node types |
| |$s\left (v_{i},v_{j}|\rho \right )$| | The proximity between node |$v_{i}$| and node |$v_{j}$| in path instance |$\rho $| |
| |$\mathcal{S}=\left \lbrace s_{i}\right \rbrace _{n}^{i=1}$| | The herb-protein transfer probability matrix and the input data for the deep autoencoder |
| |$K$| | The number of layers of the deep autoencoder |
| |$\mathcal{W}^{k}, \hat{\mathcal{W}}^{k}$| | The |$k$|-th layer weight matrix of the deep autoencoder |
| |$b^{k}, \hat{b}^{k}$| | The |$k$|-th layer biases |
| |$\boldsymbol{y}_{i}^{k}$| | The |$k$|-th layer hidden representation for |$v_{i}$| |
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.
