MathLabs

Problem 6

Let a1,a2,a3,…a_1,a_2,a_3,\ldots be a sequence of positive real numbers. Suppose that for some positive integer ss, we have an=max⁡{ak+an−k∣1≤k≤n−1}a_n=\max\{a_k+a_{n-k}\mid 1\le k\le n-1\} for all n>sn>s. Prove that there exist positive integers ℓ\ell and NN, with ℓ≤s\ell\le s, such that an=aℓ+an−ℓa_n=a_\ell+a_{n-\ell} for all n≥Nn\ge N.
Step 1 of 9: Start from the max recurrence
an=max⁡1≤k≤n−1(ak+an−k)(n>s)a_n=\max_{1\le k\le n-1}(a_k+a_{n-k})\quad(n>s)
Detailed analysis

The hypothesis is an=max⁡1≤k≤n−1(ak+an−k)(n>s)a_n=\max_{1\le k\le n-1}(a_k+a_{n-k})\quad(n>s). Whenever a summand on the right has an index greater than ss, expand that term again using the same rule.