The Cost of Parallelizing Boosting
This addresses a theoretical bottleneck for machine learning practitioners seeking efficient parallel implementations of boosting algorithms, providing foundational insights into inherent trade-offs.
The paper tackles the problem of parallelizing boosting algorithms by proving a tight lower bound that shows even slight parallelization requires an exponential blow-up in training complexity, specifically either Ω(1/γ²) rounds or an exp(d) blow-up, and complements this with an algorithm demonstrating a trade-off between parallelism and total work.
We study the cost of parallelizing weak-to-strong boosting algorithms for learning, following the recent work of Karbasi and Larsen. Our main results are two-fold: - First, we prove a tight lower bound, showing that even "slight" parallelization of boosting requires an exponential blow-up in the complexity of training. Specifically, let $γ$ be the weak learner's advantage over random guessing. The famous \textsc{AdaBoost} algorithm produces an accurate hypothesis by interacting with the weak learner for $\tilde{O}(1 / γ^2)$ rounds where each round runs in polynomial time. Karbasi and Larsen showed that "significant" parallelization must incur exponential blow-up: Any boosting algorithm either interacts with the weak learner for $Ω(1 / γ)$ rounds or incurs an $\exp(d / γ)$ blow-up in the complexity of training, where $d$ is the VC dimension of the hypothesis class. We close the gap by showing that any boosting algorithm either has $Ω(1 / γ^2)$ rounds of interaction or incurs a smaller exponential blow-up of $\exp(d)$. -Complementing our lower bound, we show that there exists a boosting algorithm using $\tilde{O}(1/(t γ^2))$ rounds, and only suffer a blow-up of $\exp(d \cdot t^2)$. Plugging in $t = ω(1)$, this shows that the smaller blow-up in our lower bound is tight. More interestingly, this provides the first trade-off between the parallelism and the total work required for boosting.