Sharp Turn in Logistic Regression

In summary, the conversation discussed the name and characteristics of a situation in logistic regression where all data points beyond a certain point are either all successes or all fails. They mentioned the use of a sigmoid function and asked about the term "saturation" in relation to this situation.
  • #1
WWGD
Science Advisor
Gold Member
7,010
10,469
Hi
I am trying to remember the name of the situation in logistic regression when all data points beyond a fixed one are all successes or all fails. So we have data points## ( a_{i1}, a_{i2},.., a_{in} , 0/1) ##, with data points ##a_{ij}##ordered; last input a Boolean and a fixed value for j after which all inputs are 0 or all are 1s. Informally, the logistic curve has to do a sharp turn when all Boolean values are equal beyond a certain point. Thanks.
 
Last edited by a moderator:
Physics news on Phys.org
  • #3
A sigmoid, thanks. Will check the link.
 
  • #4
Does the term saturation apply?
 
  • Like
Likes WWGD
  • #5
Klystron said:
Does the term saturation apply?
It was something like that, thanks.
 
  • Like
Likes Klystron

1. What is a sharp turn in logistic regression?

A sharp turn in logistic regression refers to a sudden change or discontinuity in the relationship between the independent variables and the outcome variable. It can occur when there is a non-linear relationship between the variables, or when there is a sudden shift in the data.

2. How does a sharp turn affect the logistic regression model?

A sharp turn can significantly impact the performance of a logistic regression model. It can lead to instability and inconsistency in the estimated coefficients, making it difficult to interpret the results. It can also result in overfitting, where the model fits the training data too closely and does not generalize well to new data.

3. What are the possible causes of a sharp turn in logistic regression?

There are several possible causes of a sharp turn in logistic regression. It could be due to a non-linear relationship between the variables, outliers in the data, or the presence of highly correlated variables. It could also be a result of data collection errors or an inadequate sample size.

4. How can a sharp turn be detected in logistic regression?

A sharp turn can be detected by visually inspecting the relationship between the independent variables and the outcome variable using scatter plots or other graphical methods. It can also be identified by examining the coefficients and their standard errors in the logistic regression model.

5. How can a sharp turn be addressed in logistic regression?

To address a sharp turn in logistic regression, one can try transforming the variables using techniques such as log transformation, square root transformation, or polynomial transformation. It is also important to carefully examine the data for outliers and influential points and remove them if necessary. Additionally, including interaction terms or using a different modeling approach, such as decision trees or neural networks, may also help to address a sharp turn.

Similar threads

I Are there Issues with Separation of Values in Ordinal Logistic Regression
    • Set Theory, Logic, Probability, Statistics
Replies
4
Views
1K
I Lack of Fit in Ordinal Regression -- Analysis/Alternatives?
    • Set Theory, Logic, Probability, Statistics
Replies
21
Views
2K
I Can I get thresholds from logistic regression coefficients?
    • Set Theory, Logic, Probability, Statistics
Replies
12
Views
1K
Linear Regression with Many y for each x
    • Set Theory, Logic, Probability, Statistics
Replies
2
Views
2K
A Leaving out a confounder in longitudinal regression?
    • Set Theory, Logic, Probability, Statistics
Replies
2
Views
1K
Regression Analysis of Tidal Phases
    • Set Theory, Logic, Probability, Statistics
Replies
2
Views
2K
I Is it better to use a regression or a classification?
    • Set Theory, Logic, Probability, Statistics
Replies
13
Views
1K
A Linear Regression with Non Linear Basis Functions
    • Set Theory, Logic, Probability, Statistics
Replies
4
Views
1K
How to - regression of noisy titration curve
    • Set Theory, Logic, Probability, Statistics
Replies
8
Views
2K
A The Blessings of Dimensionality and Hilbert's 6th Problem
    • Set Theory, Logic, Probability, Statistics
Replies
4
Views
1K

Hot Threads

Recent Insights

Back
Top