Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{201472,
        author = {E.Swathi and D.Vamsi Krishna and M.Vishwa Adithya},
        title = {Enhancing Medium-Term Stock Market Forecasts with Adaptive Feature Selection and Trend Dynamics},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {4963-4968},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=201472},
        abstract = {The stock market is very volatile and the change in prices within a short time period is usually full of noise, news events and high frequency algorithmic trades, making it even more hard to make quality predictions on the part of the average investor. The majority of current predictive models are based on a 1–5-day time horizon and use exclusively technical indicators, which is bound to produce unstable signals and big losses, particularly when the underlying asset is in a weak state. To respond to this, our project aims to cover 70 days of the medium- term horizon, as a result of a synergistic approach to basic stock analysis and more sophisticated methods of machine learning. Our approach involves computing a Stock Health Index (SHI) to select assets with a healthy financial base, denoising market prices with a smooth-optimal LOWESS filter and modeling trend dynamics and key factors through a dynamic feature selection process. We use a combination of machine learning models (Artificial Neural Networks, Support Vector Machines, K- Nearest Neighbors and Random Forest) and a weighted average to produce strong and actionable investment projections. Our full architecture is introduced in this paper and we show how our ensemble approach can provide greater stability and accuracy compared to using single-model predictions.},
        keywords = {Medium-Term Forecasting, Stock health index, LOWESS Smoothing, Ensemble Learning, Adaptive Feature Selection, Market Trend Dynamics.},
        month = {May},
        }