Machine Learning Basics: Confusion Matrix & Precision/Recall Simplified | By Dr. Ry @Stemplicity

This tutorial covers the basics of confusion matrix which is used to describe the performance of classification models.

The tutorial will also cover the difference between True Positives, True Negatives, False Positives, and False Negatives which can be described as follows:

• True positives (TP): cases when classifier predicted TRUE (they have the disease), and correct class was TRUE (patient has disease).

• True negatives (TN): cases when model predicted FALSE (no disease), and correct class was FALSE (patient do not have disease).

• False positives (FP) (Type I error): classifier predicted TRUE, but correct class was FALSE (patient did not have disease).

• False negatives (FN) (Type II error): classifier predicted FALSE (patient do not have disease), but they actually do have the disease

The tutorial will also cover the difference between classification accuracy, error rate, precision and recall. These metrics can be summarized as shown below:

• Classification Accuracy = (TP+TN) / (TP + TN + FP + FN)
• Misclassification rate (Error Rate) = (FP + FN) / (TP + TN + FP + FN)
• Precision = TP/Total TRUE Predictions = TP/ (TP+FP) (When model predicted TRUE class, how often was it right?)
• Recall = TP/ Actual TRUE = TP/ (TP+FN) (when the class was actually TRUE, how often did the classifier get it right?)

If you want to learn more, here’s a link to my new machine learning Classification course on Udemy:

Here’s a link to my new machine learning regression course on Udemy:

Subscribe to my channel to get the latest updates, we will be releasing new videos on weekly basis:

[khaibaromari]

This couldn't be explained as easier as this, thanks

[Rishu_Dakshin]

Well explained with proper examples. Thanks for the video Dr.Ryan

[dr.sangramsinha]

Excellent teaching. Thank you so much.

[Officialjadenwilliams]

Beautiful explanation. Great example as well!

[anujonthemove]

This is the best one I have found so far.

[Jdjdhsgxuxu]

I almost cried when I didn’t understand this. All I needed was weed and this video. Thanks.

[piyushganar]

Great explanation Dr. Ryan. i took your course on udemy and absolutely loved it. You are a great teacher :)

[xpertstrategist]

Well explained. Thanks so much for this explanation. I used it in my final AI thesis project.

[ambikashetty]

Thanks a lot sir.am confused in class.but now I clearly understood

[arjundev]

Hi,

Content is good. however i feel at 2:46 there seems to be incorrect.. when we are considering + as cancer(1) and - as no cancer(0). ideally that is 1-0 which should be False Negatives and 0-1 which should be False Positive.. please correct me if i am wrong.

[javier-medel]

An amazing explanation. Thanks for clarify the concepts

[sidharthmanne]

You have interchanged Type 1 error and Type 2 error. Type 1 error is more serious. Hence, you control is by providing an error tolerance denoted by alpha

[Holasticlogger]

This was a very nice video to recall Confusion Matrix topic. Please add AUC ROC curve topics explanation.

[nkechiesomonu]

pls i will like to join your course in udemy "under the hood" but cant find it

[wanys]

i think you mean the actual true positive when you describe the recall not just the actual true its great explain but this part may be confusing

[liornisimov]

Can I assume that the precision measure is less informative because of the small numbers in the TP & FP?

[Tommy-zbsi]

Any paper that can use as literature for perfomance measures > confusion matrix?

[sachindubey]

Hey dear i also watched your course on udemy on MATLAB, the lectures of superb i hope this content is also good

[nkristianschmidt]

Recall is a weird word to use. Recall = Sensitivity. Sensitivity is also a weird word to use. But text books are text books.