Computer Science
8 years ago

Revolutionizing Direct Mail: SVMs Outperform Traditional Algorithms for Targeted Marketing

DMIN
Comparing the Results of Support Vector Machines with Traditional Data Mining Algorithms.
Scott Pion, Lutz Hamel

Paper Summary

Related papers

Theme background image
Computer Science
4 years ago

AI Regression Techniques Unlock Powerful Predictive Insights for Real-World Challenges

Learning with Noisy Labels
Data Mining Techniques
Dimensionality Reduction
2 views
Theme background image
Computer Science
5 years ago

Decision tree outperforms logistic regression in complex datasets, improving accuracy.

Imbalanced Data
Rough Sets
Data Mining Techniques
1 view
Theme background image
Computer Science
5 years ago

New Study Reveals Best Method for Predicting Youth Internet Access

Data Mining
Dimensionality Reduction
Multivariate Analysis

The study compared different methods to predict who would respond to a life insurance mailing using direct mail data. They tested support vector machines (SVMs), logistic regression, and data mining algorithms. Results showed that SVMs performed slightly better than logistic regression, which performed better than the data mining tool. SVMs were found to be useful for handling unbalanced and noisy data, and for ranking observations based on likelihood.