International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 06 | Jun 2024
www.irjet.net
p-ISSN: 2395-0072
Regressing metrics using advance machine learning Sai Ujwal V1 1Student, Computer science and engineering, Jain University, Bangalore
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract
-
Advanced Machine Learning (AML) technologies have transformed predictive modeling, introducing sophisticated algorithms and methods to address complex regression challenges. Within AML, evaluating regression models is crucial for guiding model selection, optimization, and interpretation. The theory behind regression metrics in advanced machine learning is explored in this paper, along with the statistical characteristics, mathematical foundations, and real-world applications of important metrics like AME,SRME,SME, and squared-R (R²S). The goal of the review is to give practitioners a comprehensive understanding of these core ideas by helping to clarify them. Additionally, advanced topics such as robust regression metrics, ensemble-based evaluation, and the relationship between regression metrics and model interpretability are explored, offering insights into the latest developments in the field. Through this comprehensive analysis, the paper aims to equip researchers, practitioners, and enthusiasts in the AML community with the knowledge to effectively use regression metrics in their work.
theoretical underpinnings of regression metrics within advanced machine learning.
Key Words: Machine Learning, Statistics, Squared Mean Error, Absolute Mean Error, Standardized Root Mean Error, AML Community, Predictive Modeling.
where (n) is the number of observations, (y_i) is the target variable's actual value for observation (i), and (hat{y}_i) is the corresponding forecast value. SME is sensitive to outliers and big errors since it penalizes greater deviations more severely by squaring the errors. Finding the regression model's parameters that maximize the model's fit to the training set of data is equivalent to minimizing the mean square error. But it's important to understand that SME might not always match the goals of the modelling work, particularly in situations when the data distribution is skewed or there are a lot of outliers.
1.1 Mathematical Foundations of Regression Metrics Regression metrics quantify the discrepancy between the predicted values generated by a regression model and the actual values observed in the dataset. Understanding the mathematical underpinnings of these metrics is crucial for interpreting their significance and utility in model evaluation.
1.2 Squared mean Error (SME) Squared mean Error (SME)is a commonly used regression statistic that calculates the average squared difference between values that were predicted and those that were observed. SME has the following mathematical definition: [ text{sme} hat{y}_i)^2]
1.INTRODUCTION This Advanced Machine Learning (AML) Technologies: Predictive Modelling Enters a New Era. Advanced Machine Learning (AML) A new era of predictive modelling has been made possible by technological advancements. The creation of highly accurate and scalable solutions across various applications. In the domain of AML, regression tasks, which involve predicting continuous numerical values, are ubiquitous. Whether forecasting stock prices, estimating housing values, or predicting customer lifetime value, regression models play a central role in numerous fields.
|
Impact Factor value: 8.226
frac{1}{n}
sum_{i=1}^{n}
(y_i
-
1.3 Root Mean Squared Error (RMSE) Error Squared Root Mean (ESRM)is a variant of SME that provides a measure of prediction error in the same units as the target variable.
Despite advancements in algorithms and methodologies, evaluating regression models remains a cornerstone of the model development process. Through the use of regression metrics, practitioners can evaluate the precision, resilience, and generalizability of their models. Regression metrics quantify model performance. In-depth information about the mathematical concepts, statistical characteristics, and real-world applications of important metrics are provided in this paper's exploration of the
© 2024, IRJET
=
By taking the square root of SME, SRMEoffers a more interpretable metric that facilitates comparisons across different datasets and domains. Mathematically, SRMW is expressed as: [ text{SRME} = sqrt{text{SME}} sum_{i=1}^{n} (y_i - hat{y}_i)^2} ]
|
ISO 9001:2008 Certified Journal
=
sqrt{frac{1}{n}
|
Page 1047