Research Article
[Retracted] Recommendation Model of Tourist Attractions Based on Deep Learning
Table 1
Classification of recommendation models based on reinforcement learning.
| Type | Traditional reinforcement learning recommendation | Deep reinforcement learning recommendation | MAB-based recommendation | MDP-based recommendation | Recommendation based on value function DRL | Recommendation based on policy gradient DRL |
| Enter | Actions/States and actions | State and action/State | State and action/State | State and action/State | Output | Reward value | Q value | Q value | Probability of taking an action | Feature | Maximize step reward | Maximize total reward | Maximize total reward | Maximize total reward | Literature model | MusicCN-Bandit | Constrained PSRL | DRN | LEAP | DeepPage | DCdrift | Multi_With | Robust | CRS | DeepChain | CoLin | Bic-RL | DEERS | Top-k Corrected | RCR | Corr-Bandit | Bayes-UCB-CN | FeedRec | REINFORCE | MARDPG | MF-Bandit | ε-SVR-C | GAUM | LSIC | VL-Rec | e-TSbandit | DPG-FBE | SLATEQ | TPGR | MaHRL | DMCB | DJ-MC | Pseudo Dyna-Q | IRecGAN | HRL-Rec | POMDP-Rec | APG | Value-aware | SRL-RNN | KERL | Web-bandit | IRL-based | UDQN | LIRD | KGRL | | | GCQN | | NRRS |
|
|