Regularized Logistic Regression
Requires a Wolfram Notebook System
Interact on desktop, mobile and cloud with the free Wolfram CDF Player or other Wolfram Language products.
Requires a Wolfram Notebook System
Edit on desktop, mobile and cloud with any Wolfram Language product.
This Demonstration uses a logistic model to fit a selection of binary data. The degree of the underlying two-variable polynomial is 6. It is well known that high-degree models may lead to overfitting of the data. A common strategy to avoid overfitting is the use of regularization. This Demonstration investigates the impact of the regularization parameter 
on the final shape of the decision boundary.
Contributed by: Mikel Landajuela (November 2018)
Open content licensed under CC BY-NC-SA
Details
The logistic model (or logit model) is a widely used statistical model that, in its basic form, uses a logistic function to model a binary dependent variable.
The model is represented as:
,
with 
, a sigmoid function.
The cost function takes the form:
,
where 
is the number of training examples and 
is the regularization parameter.
The data was obtained from [1].
Reference
[1] Coursera. "Machine Learning." (Nov 29, 2018) www.coursera.org/learn/machine-learning.
Snapshots
Permanent Citation
Chains of Regular Polygons and Polyhedra
George Beck
k-Nearest Neighbor (kNN) Classifier
Ian McLeod
Multidimensional Scaling
Yaroslav Bulatov
Cluster Analysis
Jon McLoone
Cluster Analysis for 2D Points
Stephen Wolfram
Sylvester's Four-Point Problem
Ed Pegg Jr
Generating Points with Uniform Density in a Triangle, Parallelogram, or Disk
Jaime Rangel-Mondragon
B-Spline Curves from Voronoi Polygon Vertices
Enrique Zeleny
Nearest Neighbor Graph Connections
Jim Gerdy
Random Chord Paradox
Ed Pegg Jr
