AI summary by TwinMind
Lecture 10: The Completeness of the Real Numbers and Basic Properties of Infinite Series
MIT OpenCourseWare
1h 15m 36s
Cauchy Sequences Definition and Properties
- Definition: Sequence {xₙ} is Cauchy if for all ε > 0, there exists M ∈ ℕ such that for all n,k ≥ M: |xₙ - xₖ| < ε
- Intuitive meaning: if you go far enough out, any two entries are close together
- Example: xₙ = 1/n is Cauchy
- Choose M such that 1/M < ε/2
- For n,k ≥ M: |1/n - 1/k| ≤ 1/n + 1/k ≤ 2/M < ε
- Negation: {xₙ} not Cauchy if there exists ε₀ > 0 such that for all M ∈ ℕ, there exist n,k ≥ M with |xₙ - xₖ| ≥ ε₀
- Example: (-1)ⁿ is not Cauchy
- Choose ε₀ = 2
- For any M, take n = M and k = M+1
- |(-1)ᴹ - (-1)ᴹ⁺¹| = 2
Cauchy Sequences are Bounded
- Proof: For ε = 1, exists M such that for n ≥ M: |xₙ - xₘ| < 1
- Therefore |xₙ| ≤ |xₘ| + 1 for all n ≥ M
- Let B = |x₁| + |x₂| + ... + |xₘ₋₁| + |xₘ| + 1
- Then |xₙ| ≤ B for all n
Cauchy Subsequence Convergence Theorem
- If {xₙ} is Cauchy and has convergent subsequence {xₙₖ} → x, then xₙ → x
- Proof uses triangle inequality: |xₙ - x| ≤ |xₙ - xₙₖ| + |xₙₖ - x|
- First term < ε/2 (Cauchy property), second term < ε/2 (subsequence convergence)
Fundamental Theorem: Cauchy ⟺ Convergent
- Direction 1: Cauchy ⟹ Convergent
- Cauchy sequences are bounded
- By Bolzano-Weierstrass, has convergent subsequence
- By previous theorem, entire sequence converges
- Direction 2: Convergent ⟹ Cauchy
- If xₙ → x, then for ε > 0, exists M such that |xₙ - x| < ε/2 for n ≥ M
- For n,k ≥ M: |xₙ - xₖ| ≤ |xₙ - x| + |x - xₖ| < ε/2 + ε/2 = ε
Completeness of Real Numbers vs Rationals
- In ℝ: Cauchy ⟺ Convergent
- In ℚ: Convergent ⟹ Cauchy, but Cauchy ⟹ Convergent fails
- Example: Rational sequence converging to √2 is Cauchy in ℚ but doesn't converge in ℚ
- Least upper bound property is the reason for completeness in ℝ
Series Introduction
- Series ∑xₙ is formal symbol associated with sequence {xₙ}
- Partial sums: sₘ = ∑ₙ₌₁ᵐ xₙ
- Series converges if sequence of partial sums {sₘ} converges
- If convergent, series equals limit of partial sums
Series Examples
- Convergent: ∑(1/((n+1)n)) = 1
- Telescoping: 1/((n+1)n) = 1/n - 1/(n+1)
- Partial sum: sₘ = 1 - 1/(m+1) → 1
- Divergent: ∑(-1)ⁿ
- Partial sums alternate between -1 and 0
- Geometric series: ∑rⁿ (n=0 to ∞)
- Converges to 1/(1-r) if |r| < 1
- Partial sum formula: (1-rᵐ⁺¹)/(1-r)
Series Convergence Properties
- Series convergence independent of finitely many initial terms
- ∑ₙ₌₁^∞ xₙ converges ⟺ ∑ₙ₌ₘ^∞ xₙ converges for any M
Cauchy Criterion for Series
- Series is Cauchy if sequence of partial sums is Cauchy
- Series is Cauchy ⟺ Series converges (by fundamental theorem)
- Alternative characterization: For all ε > 0, exists M such that for all l > m ≥ M: |∑ₙ₌ₘ₊₁ˡ xₙ| < ε
- Intuition: tail sums become arbitrarily small
Necessary Condition for Series Convergence
- If ∑xₙ converges, then lim(n→∞) xₙ = 0
- Proof: Use Cauchy criterion with l = m to get |xₘ| < ε for large m
- Contrapositive: If lim xₙ ≠ 0 (or doesn't exist), then ∑xₙ diverges
- Application: Geometric series ∑rⁿ diverges if |r| ≥ 1 (since rⁿ ↛ 0)
- Note: This is only necessary, not sufficient condition
