Machine Learning - Data Description - Measures of Central Tendency Distribution Shapes

In this video we are discussing measures of central tendency distribution shapes

So in the previous video we have discussed what is mean, median, mode and also mid-range

So summary of the measures of the central tendency. So you want to put everything into one table

So mean, there is a sum of value divided by the total number of values

So in case of population it will be represented by mu. In case of sample, it will be represented by X bar. Median

Middle point in the data set that has been ordered. So middle point of the data set where the data set has been ordered and it will be denoted

by capital M, capital D, MD. Mode, that is the most frequent data value

It will be represented by there is no such specific symbol is given

In case of mid-range, there is the lowest value plus highest value within the data set divided

by 2 and it will be represented by MR. Now consider this respective distribution shapes

So here it is C, it is about positively skewed or right skewed

In this case what will happen this particular this particular bar will have the highest frequency So the mode will be existing here Median will be there and the mean value will be existing in this respective class

So here you can find that we have drawn histograms and this mode is that very particular

class which is having the highest frequency. That's why this is the mode here

Median will be lying in this particular class in this lower and upper limits

and mean will be existing here. So, in case of positively skewed or right skewed, the mode median

mean they will be appearing in this respective classes. Now we'll be going for the next distribution

shape that is symmetric. In that case, the middle one, the middle class is having the maximum

frequency, so the mode will be existing here along with median and mean. But in case of negatively

skewed or left skewed, you can find this one, the mode will be

be coming in this particular class and median will be there and mean will be here

So in this way we have discussed that in case of measure of central tendency, how the distribution

shapes will play a vital role and from the distribution shape we can easily guess the

respective model class or the class in which the mean and median will be existing

Thanks for watching this video