# diff b/w different models

Discussion in 'Big Data and Analytics' started by Narayana Surya, Jan 23, 2019.

1. ### Narayana Surya Well-Known Member Alumni

Joined:
Feb 27, 2018
Messages:
59
0
Hi

Can you please let me know the difference b/w different models in Data science.

below are my questions:

1)what is the difference b/w gradient boosting regressors and linear regression ,if gradient boosting regressors
gives more accuracy why you did not cover it in this course..?

2)what is the difference b/w random forest and KNN ..?
my understanding:
According to my understanding in Random forest also variables does not follow any pattern(not like in liner regression which follows normal distribution) just as KNN then why random forest called parametric test while KNN called as non parametric test

3) Also Please let me know what model i should use if i have categorical data as variables ..?

Regards,
Surya.

#1
2. ### Vikas Kumar_18 Well-Known Member Simplilearn SupportAlumni

Joined:
Dec 17, 2018
Messages:
205
37
Hi Narayan!

Both are regression algorithm. Linear regression is the standard or mother algorithm whereas gradient regression is the advanced algorithms so that it gives a better result. We have multiple algorithms but we are covering the algorithms in such a way that it would cover all mother algorithms.

The random forest also comes under nonparametric algorithms. Random forest based on Decision tree algorithms, we can say multiple collections of Decision tree called as Random forest.

KNN is a model less algorithm where each row is compared with the other all rows and then find the nearest with (k = n).

It depends upon the target column. If you have classification then go for the classification algorithms and Regression then go for the regression algorithms. If you have categorical then you need to convert it into the numeric then apply any algorithms.

I hope you got the answer.

#2