Littlestone Classes are Privately Online Learnable
This work addresses privacy-preserving online learning for machine learning practitioners, strengthening the connection between online and differentially-private learnability, though it is incremental in directly privatizing existing algorithms.
The paper tackles the problem of online classification under differential privacy constraints, showing that if the hypothesis class has constant Littlestone dimension, a private learner can achieve an expected mistake bound of O(log T), comparable to non-private optimal bounds up to a logarithmic factor, with a doubly-exponential factor for general Littlestone dimensions.
We consider the problem of online classification under a privacy constraint. In this setting a learner observes sequentially a stream of labelled examples $(x_t, y_t)$, for $1 \leq t \leq T$, and returns at each iteration $t$ a hypothesis $h_t$ which is used to predict the label of each new example $x_t$. The learner's performance is measured by her regret against a known hypothesis class $\mathcal{H}$. We require that the algorithm satisfies the following privacy constraint: the sequence $h_1, \ldots, h_T$ of hypotheses output by the algorithm needs to be an $(ε, δ)$-differentially private function of the whole input sequence $(x_1, y_1), \ldots, (x_T, y_T)$. We provide the first non-trivial regret bound for the realizable setting. Specifically, we show that if the class $\mathcal{H}$ has constant Littlestone dimension then, given an oblivious sequence of labelled examples, there is a private learner that makes in expectation at most $O(\log T)$ mistakes -- comparable to the optimal mistake bound in the non-private case, up to a logarithmic factor. Moreover, for general values of the Littlestone dimension $d$, the same mistake bound holds but with a doubly-exponential in $d$ factor. A recent line of work has demonstrated a strong connection between classes that are online learnable and those that are differentially-private learnable. Our results strengthen this connection and show that an online learning algorithm can in fact be directly privatized (in the realizable setting). We also discuss an adaptive setting and provide a sublinear regret bound of $O(\sqrt{T})$.