• retagged by
131 views
0 0 votes

Which model do you prefer, assuming both have zero training error?
Model structure (for both models):

$$
h_{\boldsymbol{\theta}}(x)=\theta_0+\theta_1 x+\theta_2 x^2+\theta_3 x^3+\theta_4 x^4+\theta_5 x^5+\theta_6 x^6+\theta_7 x^7+\theta_8 x^8
$$

Model parameters:

$$
\boldsymbol{\theta}=\left[\begin{array}{lllllll}
\theta_0, & \theta_1, & \theta_2, & \theta_3, & \theta_4, & \theta_5, & \theta_6,
&\theta_7, & \theta_8
\end{array}\right]^T
$$

$\text { A. } \quad \boldsymbol{\theta}_A=[-190.0,-135.0,310.0,45.0,-62.0,90.0,-82.0,-40.0,29.0]^T$

$\text { B. } \quad \boldsymbol{\theta}_B=\left[\begin{array}{lllllllll}
25.5, & -6.4, & -0.8, & 0.0, & 6.6, & -4.4, & 0.2, & -2.9, & 0.1
\end{array}\right]^T$

 

 

The violet line reprents model A and green line represents model B.

(Please enter 0 for A and 1 for B)

1 Answer

0 0 votes
The model B is more stable and will not overfit because it has smaller coefficients compared to model A.
Answer:
Position:
Show:

Related questions

0 0 votes
1 1 answer
129
129 views
GO Classes asked Mar 13, 2025
129 views
Good practices to avoid overfitting include:  Using a two part cost function which includes a regularizer to penalize model complexity.Using a good optimizer to minimize ...
0 0 votes
0 0 answers
131
131 views
GO Classes asked Mar 13, 2025
131 views
If the model resulting from Ridge regression is currently overfitting, what are possible things to reduce overfitting ? Select all that apply. Collect new data to increas...
0 0 votes
0 0 answers
113
113 views
GO Classes asked Mar 13, 2025
113 views
For ridge regression, if the regularization parameter is too large, which of the following is true ? Large coefficients will not be penalizedThe model will underfit the d...
0 0 votes
0 0 answers
121
121 views
GO Classes asked Mar 13, 2025
121 views
To prevent overfitting of our linear model, we would apply regularization on the model. Which of the following datasets should we evaluate using our model in order to dec...