3D Comparison Dashboard of Traditional and Open Set Threat Detection Models in Cybersecurity

"3D comparison dashboard: Left传统封闭集检测器（灰色模型误将未知木马标记为‘正常’），Right我们的开放集系统（蓝色模型正确触发‘Unknown Threat’警报），中间特征空间投影对比（传统模型的紧凑聚类 vs 开放集的分离超球面），底部性能指标：FPR从32%降至6.7%，Recall@Unknown提升89%，Neo-Tokyo全息UI风格", Create a 3D comparison dashboard showing two vehicle models. Left model is a traditional closed-set detector in gray, mislabeling unknown threats as normal. Right model is an open-set system in blue, correctly alerting for unknown threats. Feature space comparison between compact clustering of the traditional model and the separated hypersphere of the open-set model is shown in the middle. Bottom displays performance metrics where FPR drops from 32% to 6.7% and Recall@Unknown improves by 89%. Visual style reflects Neo-Tokyo aesthetics