Greedy sparsifications of sums of positive semidefinite matrices

We prove a deterministic analogue of Rudelson's sampling theorem for sums of positive semidefinite matrices. Let $A_1,\dots,A_m$ be positive semidefinite $d\times d$ matrices, and let $\lambda_1,\dots,\lambda_m \ge 0$ satisfy $$\sum_{i=1}^m \lambda_i = 1, \qquad \sum_{i=1}^m \lambda_i A_i = I_d, \qquad \|A_i\| \le M \quad\text{for all } i=1,\dots,m.$$ We show that there exists a deterministic sequence of indices $i_1,i_2,\dots \in \{1,\dots,m\}$ such that for every integer $k \ge 1$, $$\left\| \frac{1}{k}\sum_{r=1}^k A_{i_r} - I_d \right\| \le \begin{cases} \displaystyle \frac{2M\ln(2d)}{k}, & \text{if } k \le M\ln(2d),\\[2ex] \displaystyle 3\sqrt{\frac{M\ln(2d)}{k}}, & \text{if } k > M\ln(2d). \end{cases}$$ In particular, if $0<\varepsilon\le 1$ and $N \ge 9M\ln(2d)\varepsilon^{-2}$, then one can choose indices $i_1,\dots,i_N \in \{1,\dots,m\}$ such that $$\left\| \frac{1}{N}\sum_{r=1}^N A_{i_r} - I_d \right\| \le \varepsilon.$$