How to solve a problem through Machine Learning



hello and welcome my name is Susan Mansour I'm CISO certified and I kept affiliate in recent past you must have heard a lot about machine learning in my video I'm going to provide you an intuition about what machine learning is and how it is applied to the problem in its simplest form machine learning takes input and and verify it or analyze it with the output let's say we have some inputs and then we have output or outputs machine learning algorithm identify the relationship between various inputs to either one output or multiple output by analyzing this provoke performing this process machine learning model provides us an algorithm so this area will provide us some machine learning model that we can apply it to our problem to find a solution now let's see how it is what it works on our current problem here you can see in the Baloo school rectangle we have three enter guns one Pentagon with one missing value this missing value is actually our problem and our objective is to find the missing value let us name all these areas the missing value we will term it as Z all the data in this blue rectangle will be called as a data set pentagon on the right side will be our variable x2 and the left side will be variable x1 these two x we will X as inputs we will take and the Y as outputs here this Pentagon is our Y variable which has the missing value now we will plot this data in a sequence like x1 is equal to 6 4 2 7 6 4 2 7 and from x2 input we will take 5 1 3 2 next week we will put all the y-values we already know that is eleven five five nine when we try to correlate these inputs with this output we will find a model that we that will be provided by the machine learning algorithm that is y is equal to x1 plus x2 here we can verify that x + 5 6 plus 5 is 11 4 plus 1 is 5 2 plus 3 is 5 7 plus 2 is 9 when we get this equation we can use this equation which has been provided to us by the machine learning algorithm and we can use this model for finding the missing values here we will plot Z as 1/1 is we already know what is 1 and which has been provided to us 1 and then 1 plus 9 so Z is our the value of those that is 10 here this is answer in our business context this all variable could be input variable could be our customers behavior customer dynamics and the wife would be the preference to our product and in more complex way this could be our pricing the similar way we find the pricing the beauty of the machine learning algorithm is that it does not remain up to two variables machine learning algorithm could take input of millions and millions of variables and with with efficiency we'll be able to provide us this equation which will be very efficient for us which will be more data driven than our normal theories visitor ease or any other personal bias intuitive things that's why you will always see that whenever the machine learning is discussed the problem about the data quality and its value and volume is always discussed the more data you will provide to the machine learning algorithm it the output or the model it will provide you will be more efficient by now I will conclude my video and thank you for listening

Leave a Reply

Your email address will not be published. Required fields are marked *