JISA 2018 preliminary results

jisa_results_all<-read.csv("results/results_default-experiment.csv")
jisa_results_per_class<-read.csv("results/results_per_subclass_default-experiment.csv")

Source code available on github (https://github.com/harpomaxx/dga-cnn)

Crosvalidation results using datataset from edna mySQL (accesed 05/02/2018)

Dataset available here

Results considering both classes Malware DGA and Normal

jisa_results_all %>% group_by(metric) %>% summarise(mean=mean(value),sd=sd(value))

jisa_results_all %>% filter(metric=="Sensitivity" | metric == "Specificity" | metric=="F1" | metric=="Balanced Accuracy") %>%
ggplot() +
  geom_boxplot(aes(x=metric,y=value,fill=metric))+
  theme_bw()+
   theme(axis.text.x = element_text(angle = 45, hjust = 1))

Results considering malware and normal subclasses

jisa_results_per_class %>% group_by(label) %>% summarise(mean=mean(recall),sd=sd(recall))

jisa_results_per_class %>%
  ggplot()+
  geom_boxplot(aes(x=label,y=recall,color=label))+
  theme_bw()+ylab("Sensitivity (AKA Recall)")+
  theme(axis.text.x = element_text(angle = 45, hjust = 1))+
  guides(color=FALSE)