New Methods of Enhancing Prediction Accuracy in Linear Models with Missing Data

TL;DR

This paper introduces new algorithms, including a Soft Weighted Prediction method, to improve prediction accuracy in linear models with missing data by optimizing the bias-variance trade-off, outperforming existing methods.

Contribution

The paper presents novel algorithms tailored for missing data scenarios, demonstrating improved prediction accuracy through controlled over-fitting and optimized bias-variance trade-off.

Findings

SWP algorithm outperforms previous methods in non-missing scenarios

Modified algorithms improve test set MSE in missing data cases

Simulation results validate the effectiveness of the proposed heuristics

Abstract

In this paper, prediction for linear systems with missing information is investigated. New methods are introduced to improve the Mean Squared Error (MSE) on the test set in comparison to state-of-the-art methods, through appropriate tuning of Bias-Variance trade-off. First, the use of proposed Soft Weighted Prediction (SWP) algorithm and its efficacy are depicted and compared to previous works for non-missing scenarios. The algorithm is then modified and optimized for missing scenarios. It is shown that controlled over-fitting by suggested algorithms will improve prediction accuracy in various cases. Simulation results approve our heuristics in enhancing the prediction accuracy.