International Journal of Innovative Computer Science and IT Research
E-ISSN: 3067-1108
A Widely Indexed Open Access Peer Reviewed Multidisciplinary Monthly Scholarly International Journal
Plagiarism is checked by the leading plagiarism checker
Call for Paper
Volume 2 Issue 6
November-December 2026
Automated Feature Engineering for High-Dimensional Machine Learning Applications
| Author(s) | Alicja Gosiewska |
|---|---|
| Country | Poland |
| Abstract | High-dimensional machine learning applications frequently involve datasets containing thousands or millions of variables, creating substantial challenges for data representation, computational efficiency, model training, and predictive reliability. Conventional feature engineering depends heavily on domain expertise and manual experimentation, making it difficult to identify meaningful feature transformations and relevant feature subsets within large search spaces. Automated feature engineering provides a systematic approach for generating, transforming, evaluating, and selecting features with limited human intervention. This paper examines an automated feature engineering framework designed specifically for high-dimensional machine learning applications. The proposed conceptual framework integrates automated preprocessing, feature generation, feature transformation, dimensionality reduction, feature selection, search-based optimization, and model-aware evaluation into a coordinated pipeline. The methodology emphasizes hierarchical search strategies that progressively eliminate irrelevant transformations and redundant variables before computationally expensive model evaluation. The framework also incorporates statistical relevance, predictive contribution, redundancy control, computational cost, and interpretability as complementary evaluation criteria. High-dimensional feature spaces are analyzed according to their structural characteristics, including sparsity, multicollinearity, nonlinear relationships, categorical complexity, and interaction effects. The analysis demonstrates how automated feature engineering can reduce the burden associated with manual feature construction while supporting more efficient machine learning workflows. Two conceptual tables classify the principal components and optimization mechanisms associated with automated feature engineering. Particular attention is given to preventing data leakage, controlling overfitting, maintaining reproducibility, and balancing predictive performance with computational complexity. The study concludes that automated feature engineering can serve as an important component of scalable machine learning systems when combined with controlled search, robust validation, domain constraints, and transparent feature-selection mechanisms. Future developments should emphasize adaptive search, explainable feature construction, multimodal feature engineering, and integration with increasingly autonomous AutoML systems. |
| Keywords | Automated Feature Engineering, High-Dimensional Data, Machine Learning, Feature Selection, Feature Construction, AutoML, Dimensionality Reduction, Feature Transformation, Predictive Modeling, Intelligent Computing. |
| Field | Engineering |
| Published In | Volume 2, Issue 4, July-August 2026 |
| Published On | 2026-08-06 |
Share this

E-ISSN: 3067-1108
CrossRef DOI is assigned to each research paper published in our journal.
IJICSITR DOI prefix is
10.00000/IJICSITR
All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.