Deep learning in Power System Security Assessment

presented by Dixing (Dex) Xu, Linkang Du, Mingyang Sun

Training

• Input: nomial case 12k
• Output: 22 binary labels
• Model: Neural Networks (2 fully connected layers)

ax = sns.boxplot(data=[p, r, f])
ax.set(xlabel='metrics')
ax.set_xticklabels(['precision','recall','fscore'])
plt.show()

sorted_f = sorted(f)
sorted_p = sorted(p)
ax = sns.lineplot(data=[np.array(sorted_f), np.array(sorted_p)])
ax.legend(['fscore', 'precision'])
plt.show()
# print(sorted_f)
# print(sorted_p)

Thank you!