Open Access
A Framework for Adjusting Oversampling Bias in Machine Learning Models
4
Independent Researcher, Atlanta, USA
Abstract
Predictive modeling in the automotive industry often involves analyzing customer behavior to anticipate events such as vehicle purchases, service visits, or campaign responses. However, when working with imbalanced data—such as rare events like luxury vehicle purchases or high-ticket service upgrades—over-sampling techniques are commonly used. These techniques introduce bias into the sample, requiring adjustments to predicted probabilities to reflect the true population proportions. This paper explores the methodology of adjusting predicted probabilities using prior probabilities and demonstrates its application in automotive propensity models.
Keywords
Propensity modeling,
Machine Learning,
Prior Probabilities,
Oversampling Bias,
Adjusted Probability,
Automotive
References
BMC Medical Research Methodology, “Oversampling and replacement strategies in propensity score matching.” [Online]. Available: https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-021-01454-z
M. Widmann and A. Roccato, “From modeling to scoring: Correcting predicted class probabilities in imbalanced datasets.” [Online]. Available: https://www.dataversity.net/from-modeling-to-scoring-correcting-predicted-class-probabilities-in-imbalanced-datasets/
S. Rose, “Consistent estimation of propensity score functions with oversampled exposed subjects,” 2018, arXiv:1805.07684. [Online]. Available: https://arxiv.org/abs/1805.07684
K. S. Sarma, Predictive Modeling Using SAS Enterprise Miner. Cary, NC: SAS Institute Inc., 2013
G. King and L. Zeng, “Logistic regression in rare events data,” *Political Analysis*, vol. 9, no. 2, pp. 137–163, 2001. doi: 10.1093/oxfordjournals.pan.a004868.
SAS Communities, “Why do you require adjusted probability after oversampling?” 2021. [Online]. Available: https://communities.sas.com/t5/SAS-Data-Science/Why-do-you-require-adjust-probability-after-over-sampling/td-p/752224
V. Tummalapalli, “Adjusting Propensity Model Scores During Economic Shifts: A Framework for Short-Term and Long-Term Adaptation”, IJAIBDCMS, vol. 6, no. 4, pp. 247–250, Dec. 2025, doi: 10.63282/3050-9416.IJAIBDCMS-V6I4P129.
Most read articles by the same author(s)
- Vaibhav Tummalapalli, Cohort-Based Segmentation Framework for Machine Learning: Structuring Temporal Data for Enhanced Feature Engineering , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 03 (2026): Volume 03 Issue 03
Similar Articles
- Ripunjay Kumar, Sentiment Analysis of Social Network Comments for Identifying Opinion Leaders Using Machine Learning , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 06 (2026): Volume 03 Issue 06
- Prof. Jiao L. Shen, Kwa Kai Ming, A Hybrid Sentiment-Aware Machine Learning Framework for Real-Time Dynamic Pricing in E-Commerce. , International Journal of Intelligent Data and Machine Learning: Vol. 2 No. 11 (2025): Volume 02 Issue 11
- Isabella Rossi, Elena Petrova, LEVERAGING QUANTUM CONVOLUTIONAL LAYERS FOR ENHANCED IMAGE CLASSIFICATION: AN EXAMINATION OF QUANVOLUTIONAL NEURAL NETWORK CHARACTERISTICS , International Journal of Intelligent Data and Machine Learning: Vol. 2 No. 06 (2025): Volume 02 Issue 06
- Mr. Mohit Sahu, An Efficient Deep Learning Framework Model for High-Accuracy Image Visual Classification , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 07 (2026): Volume 03 Issue 07
- Dr. Jone Vakalalabure Tawake, Autonomous Optimization Method for Strengthening Transaction Completion Efficiency in Industrial Financing Operations , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 03 (2026): Volume 03 Issue 03
- Dr. Nguyen Minh Anh, Dr. Tran Hoang Nam, A Scalable Multi-Tenant Framework for AI-Driven Big Data Lake Management and Processing , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 08 (2026): Volume 03 Issue 08
- Yuki Nakamura, Isabella Romano, HYBRID DEEP LEARNING FOR TEXT CLASSIFICATION: INTEGRATING BIDIRECTIONAL GATED RECURRENT UNITS WITH CONVOLUTIONAL NEURAL NETWORKS , International Journal of Intelligent Data and Machine Learning: Vol. 2 No. 04 (2025): Volume 02 Issue 04
- Dr. Samuel Moyo, OPTIMIZING ADAPTIVE NEURO-FUZZY SYSTEMS FOR ENHANCED PHISHING DETECTION , International Journal of Intelligent Data and Machine Learning: Vol. 2 No. 05 (2025): Volume 02 Issue 05
- Elias J. Vance, Clara M. Soto, High-Frequency Data Driven Network Learning for Systemic Risk Analysis in Financial Markets , International Journal of Intelligent Data and Machine Learning: Vol. 2 No. 09 (2025): Volume 02 Issue 09
- Dr. Amir Reza Khosravi, Distributed Stream Processing Models for Financial Markets: A Theoretical Investigation of Kafka-Based Infrastructure in High-Frequency Digital Finance Systems , International Journal of Intelligent Data and Machine Learning: Vol. 3 No. 06 (2026): Volume 03 Issue 06
You may also start an advanced similarity search for this article.