Skip to content

3. Recall

Don’t miss YES

It is also called Sensitivity or True Positive Rate

jitne bhi actually me Positive the unme se kitne % ko model ne sahi guess kiya

Ab actually me sahi kitne the wo kaise dekhenge?

Note

TP - Model sahi tha aur model ne response me P (Positive kaha tha) aur actually me Positive tha
TN - Model sahi tha aur model ne response me N (Negative kaha tha) aur actually me Negative tha
FP - Model galat tha aur model ne response me P (Positive kaha tha) lekin actually me Negative tha
FN - Model galat tha aur model ne response me N (Negative kaha tha) lekin actually me Positive tha

TP & FN

Ab find karna hai ki total actual me sahi the unme se kitne % ko model ne sahi me sahi kaha

To ham Model ne jitno to sahi me sahi kaha usko total actual true se devide kar denge
Total actual true = TP + FN

\[ Recall = \frac{TP}{TP + FN} \]

Summary

Recall high → “Model Positive ko Negative kam bol raha hai”
Also FN should be less and TP should be more in order to make Recall higher

  • FN kam hai matlab model ne Positive ko negative kam bola
  • FN matlab: Model ne Positive ko Negative kah diya

Note

Positive ko galti se Negative keh dena (FN) jaha jyada critical hoga waha pe Recall ka high hona achha hai
EX: Bimari Detection (Model jinko bimari hai unko jitna kam miss kare utna achha)

“Recall tells us how many actual positives are correctly identified, so it focuses on reducing false negatives.”