how the "Most Likely / Prediction" in Model Simulator calculated??
what the algorithm or the criteria were taken to find the Most Likely / Prediction. if I want to write it as science in a research paper.
Find more posts tagged with
Sort by:
1 - 3 of
31
Hi,
It is the confidence level of the underlying machine learning model. Here is some more information:
http://hunch.net/?p=317
Technically only for the Bayes classifier the confidence equals a probablity but it has a similar meaning for all other models as well. For example in an SVM, the confidence is typically computed based on the distance of the data point to the separating hyperplane.
Hope that helps,
Ingo
IngoRM
I have doubt in two points:
1: how the confidence level calculated for the model and for each label?
2: what is the best confidence level(high or low)?
and in this result below, what means on 97% and how decided that the GOD class is the most likely
I have doubt in two points:
1: how the confidence level calculated for the model and for each label?
2: what is the best confidence level(high or low)?
and in this result below, what means on 97% and how decided that the GOD class is the most likely
First of all: please stop double posting. Spamming the community is not going to get you help faster, it will only annoy people.
I would recommend to go through some data science training, our new Academy site may be useful: http://academy.rapidminer.com
This video in particular covers some of the questions you have but there is more to discover:
A book like Elements of Statistical Learning may be helpful as well:
Hope this helps,
Ingo
Sort by:
1 - 2 of
21
Hi,
It is the confidence level of the underlying machine learning model. Here is some more information:
Technically only for the Bayes classifier the confidence equals a probablity but it has a similar meaning for all other models as well. For example in an SVM, the confidence is typically computed based on the distance of the data point to the separating hyperplane.
Hope that helps,
Ingo
First of all: please stop double posting. Spamming the community is not going to get you help faster, it will only annoy people.
I would recommend to go through some data science training, our new Academy site may be useful: http://academy.rapidminer.com
This video in particular covers some of the questions you have but there is more to discover:
A book like Elements of Statistical Learning may be helpful as well:
Hope this helps,
Ingo