0
$\begingroup$

Throughout my understanding about using Machine Learning, there was always a column whose name Target or Label, that was causing us to call methods Supervised learning. Now I am dealing with not a supervised or an unsupervised procedure. I do not know what exactly is. I have some data-frames that each of which has 3 columns. The columns have a relation with each other but it does not matter. I am looking for a way to show the effect of samples on each other. In the sense that, in my topic, theoretically, it has been proved that in a specific situation, some samples are more important than the others. Now I need to show it using Machine Learning. I have generated the required data for having that situation but I do not know how to find the effect of samples on each other.? I could have shown some differences on samples but couldn't find any meaningful scenario to be used in machine learning. To clarify, I have defined a value which a computation between each column value and I would like to know in the computation of this value in a specific index, which sample are playing a more important role. Now The question is that,

**HOW TO FIND THE EFFECT OF SAMPLES ON EACH OTHER USING SIMPLE OR ADVANCED TECHNIQUES?**

To understand better let's see some sort of data here. an example of data frame is below:

S   C     E
0   0.2   1e-2
1  -0.15  2e-2
2   0.24  2e-3
3  -0.1   3e-1
4   0.3   2e-1

the third column is a combination of the first 2 columns(There is the equation of combination) Now I need to know for example for the E[2], what is the effect of c[0],c[1],c[2],c[4].

$\endgroup$
2
  • 1
    $\begingroup$ What do you mean by 'the effect of samples on each other'? What is sample here? And usually for the best advice, it will be great if you bring some data in, even dummy to demonstrate your point clearly. $\endgroup$ Feb 2 at 11:47
  • $\begingroup$ @TwinPenguins you are right and I made some edits, I hope you find it better. Thanks $\endgroup$
    – john22
    Feb 2 at 11:55
0
$\begingroup$

First things first, columns in your dataframe are called as Feature, So, The right question for you will be How to find Co-Relation between features?

As mentioned in the question, You want to find co-relation among them,

SO, What is Co-relation ?
Answer :Correlation is a statistic that measures the degree to which two variables move in relation to each other.

Example: Ice Cream Sales

The local ice cream shop keeps track of how much ice cream they sell versus the temperature on that day, here are their figures for the last 12 days:

enter image description here

And here is the same data as a Scatter Plot:

enter image description here We can easily see that warmer weather and higher sales go together. This kind of relation is called as a positive co-relation which means if one variable increase the other gets increased too hence both of the Variables are providing the same information. In general case, it is better to drop one of them.

$\endgroup$
1
  • $\begingroup$ I know what you have written here. But I had not asked about the correlation between features as I am familiar with them. I need to know for example, T[18.1],T[25.1],T[19.4],T[22.1],T[18.5] on the amount Ice cream selling when the temperature is 23.4 $\endgroup$
    – john22
    Feb 2 at 14:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.