Average
| Method | NDCG@5 | NDCG@10 | ERR－IA@5 | ERR－IA@10 |
|---|---|---|---|---|
| Policy Gradient | 0.4164 | 0.4558 | 0.3257 | 0.3461 |
| Q Learning | 0.3997 | 0.4326 | 0.3172 | 0.3346 |
| With Baseline | 0.4182 | 0.4576 | 0.3314 | 0.3469 |
| Actor Critic | 0.3607 | 0.4089 | 0.2849 | 0.3085 |
| M2DIV | 0.4424 | 0.4852 | 0.3459 | 0.3686 |
Policy Gradient
| Data | alpha-NDCG@5 | alpha-NDCG@10 | ERR－IA@5 | ERR－IA@10 |
|---|---|---|---|---|
| folder 1 | 0.4086 | 0.4615 | 0.3212 | 0.3454 |
| folder 2 | 0.4297 | 0.4614 | 0.3275 | 0.3462 |
| folder 3 | 0.4031 | 0.4519 | 0.3209 | 0.3457 |
| folder 4 | 0.4206 | 0.4493 | 0.3218 | 0.3382 |
| folder 5 | 0.4198 | 0.4548 | 0.3370 | 0.3551 |
| average | 0.4164 | 0.4558 | 0.3257 | 0.3461 |
With Baseline
| Data | alpha-NDCG@5 | alpha-NDCG@10 | ERR－IA@5 | ERR－IA@10 |
|---|---|---|---|---|
| folder 1 | 0.4040 | 0.4900 | 0.3394 | 0.3524 |
| folder 2 | 0.4265 | 0.4637 | 0.3143 | 0.3338 |
| folder 3 | 0.4205 | 0.4521 | 0.3323 | 0.3503 |
| folder 4 | 0.4127 | 0.4361 | 0.3382 | 0.3510 |
| folder 5 | 0.4273 | 0.4461 | 0.3328 | 0.3470 |
| average | 0.4182 | 0.4576 | 0.3314 | 0.3469 |
Q Learning
| Data | alpha-NDCG@5 | alpha-NDCG@10 | ERR－IA@5 | ERR－IA@10 |
|---|---|---|---|---|
| folder 1 | 0.4080 | 0.4350 | 0.3251 | 0.3398 |
| folder 2 | 0.4026 | 0.4615 | 0.3127 | 0.3413 |
| folder 3 | 0.4082 | 0.4226 | 0.3152 | 0.3276 |
| folder 4 | 0.3890 | 0.4221 | 0.3104 | 0.3270 |
| folder 5 | 0.3905 | 0.4219 | 0.3226 | 0.3373 |
| average | 0.3997 | 0.4326 | 0.3172 | 0.3346 |
Actor Critic
| Data | alpha-NDCG@5 | alpha-NDCG@10 | ERR－IA@5 | ERR－IA@10 |
|---|---|---|---|---|
| folder 1 | 0.3522 | 0.4021 | 0.2744 | 0.2980 |
| folder 2 | 0.4023 | 0.4615 | 0.3063 | 0.3355 |
| folder 3 | 0.3147 | 0.3619 | 0.2535 | 0.2770 |
| folder 4 | 0.3299 | 0.3760 | 0.2529 | 0.2754 |
| folder 5 | 0.4046 | 0.4430 | 0.3375 | 0.3564 |
| average | 0.3607 | 0.4089 | 0.2849 | 0.3085 |