| Model . | Hyper-parameter . | HMDD v2.0 . | Searching space . | Description . |
|---|---|---|---|---|
| MTLMDA | Weight|$\_$|decay | |$3*10^{-4}$| | |$[10^{-3},10^{-4},3*10^{-4},10^{-5}]$| | |$L_{2}$| regularization coefficient. |
| Layers | 3 | |$[1,2,3,4,5]$| | Number of layers of chebGCN. | |
| Dropout | 0.3 | |$[0.1-0.9]$| | Dropout rate. | |
| |$lr$| | |$10^{-4}$| | |$[10^{-2},10^{-3},5*10^{-4},10^{-4},10^{-5}]$| | Learning rate. | |
| Projection dimension | 1024 | |$[64,128,256,512,1024,2048]$| | Node feature mapping dimension. | |
| Embedding dimension | 64 | |$[16,32,64,128,256,512]$| | Node feature embedding dimension. |
| Model . | Hyper-parameter . | HMDD v2.0 . | Searching space . | Description . |
|---|---|---|---|---|
| MTLMDA | Weight|$\_$|decay | |$3*10^{-4}$| | |$[10^{-3},10^{-4},3*10^{-4},10^{-5}]$| | |$L_{2}$| regularization coefficient. |
| Layers | 3 | |$[1,2,3,4,5]$| | Number of layers of chebGCN. | |
| Dropout | 0.3 | |$[0.1-0.9]$| | Dropout rate. | |
| |$lr$| | |$10^{-4}$| | |$[10^{-2},10^{-3},5*10^{-4},10^{-4},10^{-5}]$| | Learning rate. | |
| Projection dimension | 1024 | |$[64,128,256,512,1024,2048]$| | Node feature mapping dimension. | |
| Embedding dimension | 64 | |$[16,32,64,128,256,512]$| | Node feature embedding dimension. |
| Model . | Hyper-parameter . | HMDD v2.0 . | Searching space . | Description . |
|---|---|---|---|---|
| MTLMDA | Weight|$\_$|decay | |$3*10^{-4}$| | |$[10^{-3},10^{-4},3*10^{-4},10^{-5}]$| | |$L_{2}$| regularization coefficient. |
| Layers | 3 | |$[1,2,3,4,5]$| | Number of layers of chebGCN. | |
| Dropout | 0.3 | |$[0.1-0.9]$| | Dropout rate. | |
| |$lr$| | |$10^{-4}$| | |$[10^{-2},10^{-3},5*10^{-4},10^{-4},10^{-5}]$| | Learning rate. | |
| Projection dimension | 1024 | |$[64,128,256,512,1024,2048]$| | Node feature mapping dimension. | |
| Embedding dimension | 64 | |$[16,32,64,128,256,512]$| | Node feature embedding dimension. |
| Model . | Hyper-parameter . | HMDD v2.0 . | Searching space . | Description . |
|---|---|---|---|---|
| MTLMDA | Weight|$\_$|decay | |$3*10^{-4}$| | |$[10^{-3},10^{-4},3*10^{-4},10^{-5}]$| | |$L_{2}$| regularization coefficient. |
| Layers | 3 | |$[1,2,3,4,5]$| | Number of layers of chebGCN. | |
| Dropout | 0.3 | |$[0.1-0.9]$| | Dropout rate. | |
| |$lr$| | |$10^{-4}$| | |$[10^{-2},10^{-3},5*10^{-4},10^{-4},10^{-5}]$| | Learning rate. | |
| Projection dimension | 1024 | |$[64,128,256,512,1024,2048]$| | Node feature mapping dimension. | |
| Embedding dimension | 64 | |$[16,32,64,128,256,512]$| | Node feature embedding dimension. |
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.
