You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
<p>The evaluation method for DD-Ranking is grounded in the essence of dataset distillation, aiming to better reflect the information content of the synthesized data by assessing the following two aspects:</p>
182
184
<ol>
183
185
<li>
184
-
<p>The degree to which the original dataset is recovered under hard labels (hard label recovery): $\text{HLR} = \text{Acc}<em>{\text{full-hard}} - \text{Acc}</em>{\text{syn-hard}}$</p>
186
+
<p>The degree to which the original dataset is recovered under hard labels (hard label recovery): \( \text{HLR} = \text{Acc.} \text{full-hard} - \text{Acc.} \text{syn-hard} \)</p>
185
187
</li>
186
188
<li>
187
-
<p>The improvement over random selection when using personalized evaluation methods (improvement over random): $\text{IOR} = \text{Acc}<em>{\text{syn-any}} - \text{Acc}</em>{\text{rdm-any}}$</p>
189
+
<p>The improvement over random selection when using personalized evaluation methods (improvement over random): \( \text{IOR} = \text{Acc.} \text{syn-any} - \text{Acc.} \text{rdm-any} \)</p>
188
190
</li>
189
191
</ol>
190
-
<p>$\text{Acc.}$ is the accuracy of models trained on different samples. Samples' marks are as follows:</p>
192
+
<p>\(\text{Acc.}\) is the accuracy of models trained on different samples. Samples' marks are as follows:</p>
191
193
<ul>
192
-
<li>$\text{full-hard}$: Full dataset with hard labels;</li>
193
-
<li>$\text{syn-hard}$: Synthetic dataset with hard labels;</li>
194
-
<li>$\text{syn-any}$: Synthetic dataset with personalized evaluation methods (hard or soft labels);</li>
195
-
<li>$\text{rdm-any}$: Randomly selected dataset (under the same compression ratio) with the same personalized evaluation methods.</li>
194
+
<li>\(\text{full-hard}\): Full dataset with hard labels;</li>
195
+
<li>\(\text{syn-hard}\): Synthetic dataset with hard labels;</li>
196
+
<li>\(\text{syn-any}\): Synthetic dataset with personalized evaluation methods (hard or soft labels);</li>
197
+
<li>\(\text{rdm-any}\): Randomly selected dataset (under the same compression ratio) with the same personalized evaluation methods.</li>
196
198
</ul>
197
199
<p>To rank different methods, we combine the above two metrics as DD-Ranking Score:</p>
<p>The evaluation method for DD-Ranking is grounded in the essence of dataset distillation, aiming to better reflect the information content of the synthesized data by assessing the following two aspects:</p>
182
184
<ol>
183
185
<li>
184
-
<p>The degree to which the original dataset is recovered under hard labels (hard label recovery): $\text{HLR} = \text{Acc}<em>{\text{full-hard}} - \text{Acc}</em>{\text{syn-hard}}$</p>
186
+
<p>The degree to which the original dataset is recovered under hard labels (hard label recovery): \( \text{HLR} = \text{Acc.} \text{full-hard} - \text{Acc.} \text{syn-hard} \)</p>
185
187
</li>
186
188
<li>
187
-
<p>The improvement over random selection when using personalized evaluation methods (improvement over random): $\text{IOR} = \text{Acc}<em>{\text{syn-any}} - \text{Acc}</em>{\text{rdm-any}}$</p>
189
+
<p>The improvement over random selection when using personalized evaluation methods (improvement over random): \( \text{IOR} = \text{Acc.} \text{syn-any} - \text{Acc.} \text{rdm-any} \)</p>
188
190
</li>
189
191
</ol>
190
-
<p>$\text{Acc.}$ is the accuracy of models trained on different samples. Samples' marks are as follows:</p>
192
+
<p>\(\text{Acc.}\) is the accuracy of models trained on different samples. Samples' marks are as follows:</p>
191
193
<ul>
192
-
<li>$\text{full-hard}$: Full dataset with hard labels;</li>
193
-
<li>$\text{syn-hard}$: Synthetic dataset with hard labels;</li>
194
-
<li>$\text{syn-any}$: Synthetic dataset with personalized evaluation methods (hard or soft labels);</li>
195
-
<li>$\text{rdm-any}$: Randomly selected dataset (under the same compression ratio) with the same personalized evaluation methods.</li>
194
+
<li>\(\text{full-hard}\): Full dataset with hard labels;</li>
195
+
<li>\(\text{syn-hard}\): Synthetic dataset with hard labels;</li>
196
+
<li>\(\text{syn-any}\): Synthetic dataset with personalized evaluation methods (hard or soft labels);</li>
197
+
<li>\(\text{rdm-any}\): Randomly selected dataset (under the same compression ratio) with the same personalized evaluation methods.</li>
196
198
</ul>
197
199
<p>To rank different methods, we combine the above two metrics as DD-Ranking Score:</p>
0 commit comments