Prove the dominated convergence theorem.
- User
Prove the dominated convergence theorem.
- Assistant
Dominated Convergence Theorem
Theorem. Let be a measure space. Let be a sequence of -measurable functions such that -a.e., where is -measurable. Suppose there exists a -integrable function such that for -a.e. and all . Then:
- -a.e.,
- is -integrable (i.e., ),
- ,
- In particular, .
(The theorem holds for real- or complex-valued functions; the proof below is for real-valued, but extends straightforwardly.)
Proof
Since are measurable and converge pointwise -a.e. to , the limit function is measurable -a.e.
Step 1: -a.e. and .
By assumption, -a.e. for each , and -a.e. Thus, -a.e. (taking pointwise limits).
The functions are nonnegative and dominated by the integrable . Apply Fatou's lemma: [ \int_X |f| , d\mu = \int_X \liminf_{n \to \infty} |f_n| , d\mu \leq \liminf_{n \to \infty} \int_X |f_n| , d\mu \leq \int_X g , d\mu < \infty. ] Hence, is -integrable.Step 2: .
Define . Then -a.e., , and -a.e., where is integrable.
Now consider the nonnegative functions -a.e. Note that [ \liminf_{n \to \infty} k_n = \liminf_{n \to \infty} (2g - h_n) = 2g - \limsup_{n \to \infty} h_n = 2g \quad \mu\text{-a.e.} ] (by -a.e.). Apply Fatou's lemma to : [ \int_X \liminf_{n \to \infty} k_n , d\mu \leq \liminf_{n \to \infty} \int_X k_n , d\mu, ] which simplifies to [ \int_X 2g , d\mu \leq \liminf_{n \to \infty} \left( 2\int_X g , d\mu - \int_X h_n , d\mu \right) = 2\int_X g , d\mu - \limsup_{n \to \infty} \int_X h_n , d\mu. ] Rearranging gives [ 0 \leq -\limsup_{n \to \infty} \int_X h_n , d\mu \implies \limsup_{n \to \infty} \int_X h_n , d\mu \leq 0. ] Since for all , it follows that , i.e., [ \lim_{n \to \infty} \int_X |f_n - f| , d\mu = 0. ]Step 3: .
By the triangle inequality and Step 2, [ \left| \int_X f_n , d\mu - \int_X f , d\mu \right| \leq \int_X |f_n - f| , d\mu \to 0. ]This completes the proof. ∎
(Remark: This proof relies only on Fatou's lemma and basic measure theory. Alternative proofs use Egoroff's theorem for uniform convergence on finite-measure sets or density of simple functions in .)
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.