What is the meaning of the denominator P[E] in the 3rd video:

[Sourabh Sinha]

Is it the "joint conditional probabilities of the attributes" i.e. = P[sunny, cool, humid, windy]? Why we can't calculate it?

Thanks

[Sourabh Sinha]

I meant section 3.3 video lecture: "Using probabilities" 

[Huihui Duan]

Hi Sourabh: 

You are right, we CAN calculate P[E] = P[(sunny, cool, humid, windy) | yes]*P[yes] + P[(sunny, cool, humid, windy) | no]*P[no]. Actually, P[E] is the sum of 0.0053 and 0.0206. However, due to P[E] will be always a fixed number, we don't need to calculate it; because we will normalize the probabilities as discusses in the video.

Hope this helps. 

Best,

Huihui

[Fernando DM]

Hi Sourabh, 

Just in case, keep in mind that "sunny" is a category and not a variable, and "windy" is a variable and not a category, E (evidence) is the combination of the values for the new day (Sunny,Cool,High,True), when normalized the Pr (yes) is to get the Pr of YES, without knowing about E (then it is not considered in this point).  

***** ie (A new day) *******

Outlook Temp. Humidity Wind =>Play

Sunny   Cool  High     True =>?

Likelihood of the two classes

For “yes” = 2/9 * 3/9 * 3/9 * 3/9 * 9/14 = 0.0053

For “no” = 3/5 * 1/ * 4/5 * 3/5 * 5/14 = 0.0206

Finally:

Conversion into a probability by normalization:

P(“yes”) = 0.0053 / (0.0053 + 0.0206) = 0.205

P(“no”) = 0.0206 / (0.0053 + 0.0206) = 0.795

Since the probability of play=NO, is greater than the probability of play=YES, this new record (new day) is classified as Play = NO.

However, check out: Lesson 3.3 - page 21 and the "CHAPTER 4 Algorithms: The Basic Methods" from the course book.

For any further information, please let me know.

Best Regards,

Ing. Fernando García

Data Mining Specialist

ar.linkedin.com/in/fernand0garcia/