sbottorf9926 sbottorf9926

04-09-2017
Business

contestada

Using the concept of overfitting, explain why when a model is fit to training data, zero error with those data is not necessarily good

Respuesta :

Kalahira Kalahira

13-09-2017

Having zero error does not allow for the complexity of real-world data to be introduced in the future. Having a perfect fit of the model could lead to data that shows some sort of a skew when another set of data is tested at some point in the future.

Answer Link

Otras preguntas

Based on the excerpt, what inferences can be made about the story’s setting? it is located on a farm with many field hands and employees. It is located on a wid

help answer with explanation please

What does "Variation" mean? Two or more things are different in a population or group Two or more things arethe same in a population or group

___________ are the most common cause of on-the-job injuries to construction workers.

Which of the following results when homologous chromosomes cross over in meiosis? Two sister chromatids get tangled, resulting in one re-sequencing its DNA. DNA

Element X decays radioactively with a half life of 8 minutes. If there are 170 a grams of Element X, how long, to the nearest tenth of a minute, would it take t

On December 1, a company accepted a $5,000, 4%, 90-day note. How much accrued interest will be recorded as an adjusting entry on December 31, the end of the acc

The U.S. Supreme Court decision in Plessyv. Ferguson found the doctrine of separate but equal to be constitutional. What was one African American response in th

If my dot plot is not skewed or symmetrical which measure of variability should I use? (15 data points) PLEASE HELP THE ASSIGNMENT IS DUE TONIGHT! these are th

From the coordinate 8,12 to 10 and 15 what was the scale factor