Scribd
Upload
6 views3 pages

Lo 4

Uploaded by

Urwa Rehman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
6 views3 pages

Lo 4

Uploaded by

Urwa Rehman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Lo4, p2

1. Model Accuracy and Prediction Performance:


a. Mean Absolute Error (MAE)
Low MAE means the model's predictions are close to actual values.
High MAE means the model has a high prediction error, which could
signal poor performance.
b. Mean Squared Error (MSE) and Root Mean Squared Error (RMSE):
.MSE gives the average squared differences between predicted and
actual values. It is more sensitive to large errors due to the squaring of
differences.
.RMSE is the square root of MSE and provides a more interpretable
measure in the original units of the target variable.
.A low MSE/RMSE suggests good predictive performance.
.High MSE/RMSE means there are large prediction errors, indicating
the model isn't performing well.
c. Mean Absolute Percentage Error (MAPE):
Interpretation: MAPE provides an error metric as a percentage of the
actual values, useful for understanding the relative error.

 Low MAPE indicates that the model is generally accurate in


predicting the target variable.
 High MAPE means predictions are far from the actual values in
terms of percentage error.

2. Generalization:
a. Cross-Validation Performance:
Interpretation: Cross-validation (e.g., k-fold cross-validation) helps ensure the
model's ability to generalize well to unseen data.

 If cross-validation results show consistently good performance (low


MAE, MSE, etc.), it suggests that the model is generalizing well.
 Overfitting is a concern if the model performs very well on training
data but poorly on validation/test data.
 Underfitting occurs when the model performs poorly on both training
and test data.

b. Train/Test Split Evaluation:

Interpretation: Comparing the performance on the training data versus the test
data helps to assess if the model is overfitting or underfitting.
 Overfitting: If the model performs much better on the training data
compared to the test data, it may have memorized patterns in the
training data, making it unable to generalize to new data.
 Underfitting: If both training and test performance are poor, the model
may be too simple (e.g., linear regression may not capture the
complexity of the relationship).

3. Model Assumptions Check:

Linear regression makes several assumptions that need to be validated for the
model to be effective:

 Linearity: The relationship between predictors and the target variable


should be linear.
 Independence: The residuals should be independent, with no
autocorrelation.
 Homoscedasticity: The variance of residuals should be constant across all
levels of the independent variable(s).
 Normality: Residuals should be normally distributed (especially for
hypothesis testing and confidence intervals).

4. Conclusion: check your model is effective then write down:

Effective Model:

 High R2R^2R2 (or adjusted R2R^2R2) indicating a good fit.


 Low MAE, MSE, RMSE, and/or MAPE showing predictive accuracy.
 No significant patterns in residuals (homoscedasticity, normality).
 Good generalization performance across training and test sets (via cross-
validation or hold-out validation).
 Meets assumptions of linear regression (linearity, normality, independence,
and homoscedasticity).

Ineffective Model:

 high MAE, MSE, RMSE, or MAPE.


 Residual plots showing patterns (indicating misspecification of the
model).
 Poor generalization on test data, suggesting overfitting.
 Violations of linear regression assumptions, necessitating model
adjustments or the use of a more complex model.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy