2. Why don’t we use the sum of the absolute value of the deviations, instead of the sum of squares of the deviations?Let x1, …, xn be given, where they are all some real numbers. Define the following function: resimdeki fonksiyon http://img527.imageshack.us/my.php?image=functionyw4.jpg Can you fi
2. Why don’t we use the sum of the absolute value of the deviations, instead of the sum of squares of the deviations?
Let x1, …, xn be given, where they are all some real numbers. Define the following function:
resimdeki fonksiyon
img527.imageshack.us
Can you find an expression for y* which minimizes g(y) in terms of x1, …, xn? What are the difficulties you encounter in this process? Do you think there may be alternative minimizers (alternative optima) for g? Can you claim that the sample mean, , is a minimizer of g? Why or why not? Explain. Note that you do not need to find y*, only think about how you can find it if necessary.
Yardımcı olursanız çok sevinirim ipucu olabilecek herhangi birşeyler de yazsanız olur.Teşekkürler