Linear regression
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
1. random variable basics:
a) Given that
is normed, non-negative & linear [ specifically as in Chpt.5, p.4 ]
verify that the continuity property:
is equivalent to σ-linearity:
(b) Verify
c)
2. Lebesgue linear spaces
a) For R-valued random variables, any p > 0, we define, exactly as on p.39 of Chpt.5,
But in that case, specifically for p = 1, for any X ∈ L1, verify that |EX| ≤ E|X|, and describe the circumstances for equality.
b) For Rn-valued random variables as on pp.27-31, Chpt.6, any p > 0, we define
So explain simply how it is that
c) For any
, show that |EX| ≤ E|X| and indicate the precise circumstances for equality.
3. Linear regression in practice
a). For X1,...,Xk & Y in L2, the correlation coefficient of Y with
= op(Y |1, X1,...,Xk) is usually/often referred to in the literature on multivariate statistics, as the multiple correlation of Y on X:
In the notation of ESP6.12 show that if |Σ| > 0 and σY > 0, then
b). In the special case of polynomial regression, we find that for any X & Y in L2, there will be unique coefficients β0,..., βk such that
4. statistical independence & conditional expectation
Given any distribution
and suppose X statistically independent of Y [Chpt.6, Defn.5.7.1, p.53, Eqn(103)]
— denoted
(a) Show that
(b) For any
explain why it is automatic that
(c) Prove that
5. Indefinite integral of
with respect to P
denoted
or, much more simply, by dλ = XdP.
Verify the following :
a) λ is σ-additive.
b) uniqueness :
2025-11-12