The goal of regression model is either to minimize the prediction error or to maximize a classification accuracy depending on the type of the model. Is there any optimization algorithm that is more efficient in fitting the regression model to a training data set? especially when the input data are not correlated with the out put or does not conform with expectations of the model like in other statistical distributions. In other words What is the suitable and efficient optimization algorithm that can be use to estimate the coefficients of a regression model which can minimize the prediction error or maximize the classification accuracy as the case may be?
You can look into support vector machine based classification/regression in this regard.
It depends on many factors such as the method and dataset.
Please give more details about the project and the dataset.
Dear Umar M. Modibbo ,
Regression models are fit on training data using linear regression and local searchoptimization algorithms. Models like linear regression and logistic regression aretrained by least squares optimization, and this is the most efficient approach to finding coefficients that minimize error for these models.
https://machinelearningmastery.com/optimize-regression-models/
Regards,
Shafagat
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training