How do different data imputation techniques impact the performance of predictive analytics models in incomplete datasets?

Author(s)	Mr. Rehan Patel
Country	India
Abstract	In predictive analytics, model accuracy and reliability based on data-driven models are typically inhibited by missing values within datasets, which are common in real-world data. The current paper investigates the essential function of insinuation techniques in ensuring dataset integrity and enhanced model performance. Through categorizing missing data as MCAR (Missing Completely at Random), MAR (Missing at Random), and MNAR (Missing Not at Random), the paper identifies appropriate approaches to handle each category. Comparison of various imputation techniques—mean, median, neighboring neighbors (k-NN), regression, and advanced classifier-based approaches—is presented to evaluate their impact on model outcomes. The paper also demonstrates the impact of inappropriately handling missing data, leading to bias and reduced predictive power. A realworld healthcare scenario using the MIMIC-III dataset is provided to demonstrate improved performance of advanced methods like MICE and MissForest in handling MAR data.Overall, the research emphasizes the need to choose an imputation approach, which is commensurate with the missingness type, in the development of robust and interpretable predictive models.
Keywords	data imputation techniques, predictive analytics
Published In	Volume 7, Issue 5, September-October 2025
Published On	2025-09-02
DOI	https://doi.org/10.36948/ijfmr.2025.v07i05.54709

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research