1: \begin{abstract}%

2: Motivated by the high-frequency data streams continuously generated, real-time learning is becoming increasingly important.

3: These data streams should be processed sequentially with the property that the stream may change over time.

4: In this streaming setting, we propose techniques for minimizing a convex objective through unbiased estimates of its gradients, commonly referred to as stochastic approximation problems.

5: Our methods rely on stochastic approximation algorithms due to their computationally advantage as they only use the previous iterate as a parameter estimate.

6: The reasoning includes iterate averaging that guarantees optimal statistical efficiency under classical conditions.

7: Our non-asymptotic analysis shows accelerated convergence by selecting the learning rate according to the expected data streams.

8: We show that the average estimate converges optimally and robustly to any data stream rate.

9: In addition, noise reduction can be achieved by processing the data in a specific pattern, which is advantageous for large-scale machine learning.

10: These theoretical results are illustrated for various data streams, showing the effectiveness of the proposed algorithms.%

11: \end{abstract}

12: