Efficient Methods for Non-stationary Online Learning
Authors: Peng Zhao, Yan-Feng Xie, Lijun Zhang, Zhi-Hua Zhou
JMLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we provide empirical studies to evaluate our proposed methods. We conduct experiments on the synthetic data. |
| Researcher Affiliation | Academia | Peng Zhao EMAIL Yan-Feng Xie EMAIL Lijun Zhang EMAIL Zhi-Hua Zhou EMAIL National Key Laboratory for Novel Software Technology, Nanjing University, China School of Artificial Intelligence, Nanjing University, China |
| Pseudocode | Yes | Algorithm 1 Efficient Algorithm for Minimizing Dynamic Regret Algorithm 2 Efficient Algorithm for Problem-dependent Adaptive Regret Algorithm 3 Efficient Algorithm for Problem-dependent Interval Dynamic Regret Algorithm 4 Efficient Control Algorithm for Dynamic Policy Regret Algorithm 5 Efficient Algorithm for Adaptive Regret under PCA Setting |
| Open Source Code | No | No explicit statement about open-source code or a repository link was found in the paper. |
| Open Datasets | No | We conduct experiments on the synthetic data. We consider the following online regression problem. Let T denote the number of total rounds. At each round t [T] the learner outputs the model parameter wt W Rd and simultaneously receives a data sample (xt, yt) with xt X Rd being the feature and yt R being the corresponding label. ... To simulate the non-stationary online environments, we control the way to generate the date samples {(xt, yt)}T t=1. |
| Dataset Splits | No | For dynamic regret minimization, we simulate piecewise-stationary model drifts, as dynamic regret will be linear in T and thus vacuous when the model drift happens every round due to a linear path length measure. Concretely, we split the time horizon evenly into 25 stages and restrict the underlying model parameter w t to be stationary within a stage. For adaptive regret minimization, we simulate gradually evolving model drifts, where the underlying model parameter w t+1 is generated based on the last-round model parameter w t with an additional random walk in the feasible domain W. |
| Hardware Specification | Yes | We use a machine with a single CPU (Intel(R) Core(TM) i9-10900K CPU @ 3.70GHz) and 32GB main memory to conduct the experiments. |
| Software Dependencies | No | In the experiment, we use scipy.optimize.Nonlinear Constraint to solve it to perform the projection onto the feasible domain. (No version number for scipy is provided.) |
| Experiment Setup | Yes | In the simulations, we set T = 20000, the domain diameter as D = 6, and the dimension of the domain as d = 8. The feasible domain W is set as an ellipsoid W = w Rd | w Ew λmin(E) (D/2)2 , where E is a certain diagonal matrix and λmin(E) denotes its minimum eigenvalue. ... We repeat the experiments for five times with different random seeds and report the results (mean and standard deviation) in Figure 3. ... For dynamic regret minimization, we simulate piecewise-stationary model drifts, as dynamic regret will be linear in T and thus vacuous when the model drift happens every round due to a linear path length measure. Concretely, we split the time horizon evenly into 25 stages and restrict the underlying model parameter w t to be stationary within a stage. For adaptive regret minimization, we simulate gradually evolving model drifts, where the underlying model parameter w t+1 is generated based on the last-round model parameter w t with an additional random walk in the feasible domain W. The step size of random walk is set proportional to D/T to ensure a smooth model change. |