Validate for online integration

Hey @lis07128 ,

as we discussed together here some todos for validating our anomaly detection:

• one loss value per window
• create ROC-curve (on batch level)
• set labels globally on False and label some anomaly windows (qualitiative identification) on True
• for qualitiative identification create helper fn "labeling_bias" to synchronise traj and better isolate anomalous windows
• choose values for the threshold that represent the whole spectrum (~ some FP and no TP)
• create confusion matrix (on batch level)
• adapt the labels of the conf matrix that we have anomaly as first entry and nominal as second, leading to TP on 1 row -1 col and FP on 1-2, FN on 2-1 and TN on 2-2
• nice would be to summarize the results on some slides