Questions tagged [mse]
MSE stands for Mean Squared Error. It is a measure of the performance of an estimate or prediction, equal to the mean squared difference between the observed values and the estimated / predicted values.
439 questions
1
vote
1
answer
38
views
Average MSE over training set less or equal to average MSE over test set (with other clarifying questions)
The following is a question from Elements of Statistical Learning, followed by some questions regarding my understanding of machine learning and notation, and then lastly followed by my attempt of the ...
- 11
0
votes
0
answers
21
views
Plotting Training VS Testing Curve
I am using gradient boosting regressor from scikit-learn with squared error as the loss function. Then i want to plot the training set vs test set curve. Based on what i read, it is used to see the ...
- 271
5
votes
2
answers
364
views
Why can a model with higher MSE still have a higher R² than another model
Two mdoels trained on different training datasets (related but not exactly the same) and tested on the exact same datasets, the model that has higher MSE, still has a higher R², is that possible?
- 183
3
votes
1
answer
91
views
Lower bound for MSE, based on sample mean and variance
Short question: For two unknown samples $A$ and $B$ of size $n$, if only their sample mean and sample variances are known, what can be said about $MSE(A,B)$ ?
Long version: To be more precise, I ...
- 41
5
votes
3
answers
207
views
Comparison of candidate models obtained with possibly "non-comparable AIC": use RSS, MSE, or adjusted MSE as an alternative instead... or?
My question relates to the comparison of candidate models for which their parameter estimates have been produced with different methods and R packages.
As a fictive example, the MASS (continuous ...
- 263
1
vote
0
answers
21
views
How to further reduce MSE/RMSE in CatBoost after tuning, data augmentation, and outlier removal? [duplicate]
I am using the CatBoost model with 100 data points and I have done data augmentation, hyperparameter tuning,cross validation, added isolation forest, and randomizedsearchCV and at the end I could have ...
- 11
0
votes
0
answers
18
views
Why does my GNN-LSTM model fail to generalize with full training data for a spatiotemporal prediction task? [duplicate]
I'm working on a spatiotemporal prediction problem where I want to forecast a scalar value per spatial node over time. My data spans multiple spatial grid locations with daily observations.
Data Setup
...
- 43
2
votes
1
answer
227
views
Minimizing MSE for the estimator of population variance
The screenshot below is from the Wikipedia page on Mean Squared Error. Can someone please either
help me understand the last highlighted sentence (given the claim is correct),
or confirm my suspicion ...
- 137
2
votes
1
answer
177
views
Relationship related to sum of squared deviations
I'm trying to prove this equality, in the context of linear regression:
$$
\frac{\sum x_i^2}{n S_{xx}} = \frac{1}{n} + \frac{\bar{x}^2}{S_{xx}}
$$
where
$$
S_{xx} = \sum (x_i - \bar{x})^2
$$
My ...
- 143
7
votes
1
answer
728
views
Exploiting MSE of binary data for fast search
I have a huge database of binary vectors. Given an incoming vector, I want to find the closest vector in the database in terms of MSE and return the MSE score. So far I have been doing this search ...
- 71
1
vote
1
answer
91
views
Regarding the definition of unbiasedness and connections to MSE
My impression is that the definition of an unbiased estimator is always made for a single parameter. Is there a reason that the definition cannot be extended for a vector of parameters? For example, ...
- 147
1
vote
0
answers
51
views
Mean squared error of an estimator of samples that are uniformly distributed in (-a,a)
Let $MSE(\hat{\theta}) := \mathbb{E}((\hat{\theta}-\theta)^2)$ be the mean squared error of a statistic $\hat{\theta}$.
My question is at the end of the post. The rest is my workout.
Let $X_{1},...,...
- 111
1
vote
1
answer
374
views
Way too large MSE for random forest model
I'm currently working on building Random Forest Models in python. My topic is to investigate the Imoportance of specific variables for the accuracy of machine learning to explain the market ...
- 11
4
votes
2
answers
398
views
Deriving MSE($\hat{\beta}$) under Linear regression
I was able to derive the MSE, but there's a part of the derivation which I don't really get. Here's what I got:
Facts:
$\mathbb{E}(\hat{\beta})=\hat{\beta}\space$ (unbiased estimator)
$\text{Cov}(\...
- 391
5
votes
1
answer
241
views
On page 31 of ISL, why is the minimum possible test MSE over all methods (dashed line) 1 instead of another number?
I am reading James, Witten, Hastie and Tibshirani's Introduction to Statistical Learning (2nd ed). On page 31, specifically Figure 2.9 in the right panel, I don't understand why the minimum possible ...
