this is for holding javascript data
chengds added subsection_Fixed_2C_The_latest__.tex
almost 9 years ago
Commit id: 66e7d71361f0770f7fb714cded496d1183fc8a21
deletions | additions
diff --git a/subsection_Fixed_2C_The_latest__.tex b/subsection_Fixed_2C_The_latest__.tex
new file mode 100644
index 0000000..c25bee9
--- /dev/null
+++ b/subsection_Fixed_2C_The_latest__.tex
...
\subsection{Fixed 2C}
The latest batch of 2C experiments have corrected some errors in the prototxt files (traits 7 to 10 were affected) and completed the missing ones (traits 1 to 4). Moreover, in preliminary experiments, we found out that it was possible to reach near 100\% training accuracy by fixing the first 4 conv layers, and letting only the last 3 fully connected layers (InnerProduct ones) fine-tune on the PsychoFlickr data. Unfortunately, the training tended to overfit, showing a test accuracy graph with an "elbow" path. To reduce overfitting, we increased the weight decay parameter dramatically in order to avoid any "neuron specialization" (neural units with high weights values), but this limited the max training accuracy (not needed anyway). Together, fixing the conv layers and raising weight decay, we achieved the best 2C results so far. See Figure 1, where the self traits are still unimpressive, but the attributed ones clearly show predictions above chance.
