Confusion Matrix is indeed confusing at first, but after taking a deep glance at it and figuring out what those FP, FN words really convey, it's really easy!
Great video!! BTW I hate the TruePositive, FalsePositive etc, taxonomy because It's so CONFUSING (heh), I prefer to use False alarm(FalsePositive), and Missed (or Missed defect, or something in this way - FalseNegative), because I can much easier imagine the Red alarm going insane in some building "sir rockets are coming" and then someone just says "oooh its just false alarm". And I can easily imagine another similiar situation for the Missed example..
Confusion Matrix is indeed confusing at first, but after taking a deep glance at it and figuring out what those FP, FN words really convey, it's really easy!
Great video!! BTW I hate the TruePositive, FalsePositive etc, taxonomy because It's so CONFUSING (heh), I prefer to use False alarm(FalsePositive), and Missed (or Missed defect, or something in this way - FalseNegative), because I can much easier imagine the Red alarm going insane in some building "sir rockets are coming" and then someone just says "oooh its just false alarm". And I can easily imagine another similiar situation for the Missed example..
Totally agree. After several years in data science I still need to spend an embarrassing amount of time understanding what a false positive means
It seems to me that a confusion matrix is very similar to a Chi-Square result? My prof had us summarize Chi-Square with table as well.
As always, it is another great video. I would like to hear from you about the multi-class classification model assessments.
Thanks for the explanation.
You're welcome!
Thanks for the awesome content. I could not get quit well the Recall. Please explain a bit more here. Thanks
#confusion#matrix#machinelearning#deep#precision#recall F1 #score#accuracy#true#positive #negative!
ruclips.net/video/YlFgsaxagX0/видео.html