Benefiting from high capacity for capturing complex temporal patterns, deep learning (DL) has significantly advanced time series forecasting (TSF). However, deep models tend to suffer from severe overfitting due to the inherent vulnerability of time series to noise and anomalies. The prevailing DL paradigm uniformly optimizes all timesteps through the MSE loss and learns those uncertain and anomalous timesteps without difference, ultimately resulting in overfitting. To address this, we propose a novel selective learning strategy for deep TSF. Specifically, selective learning screens a subset of the whole timesteps to calculate the MSE loss in optimization, guiding the model to focus on generalizable timesteps while disregarding non-generalizable ones. Our framework introduces a dual-mask mechanism to target timesteps: (1) an uncertainty mask leveraging residual entropy to filter uncertain timesteps, and (2) an anomaly mask employing residual lower bound estimation to exclude anomalous timesteps. Extensive experiments across eight real-world datasets demonstrate that selective learning can significantly improve the predictive performance for typical state-of-the-art deep models, including 37.4% MSE reduction for Informer, 8.4% for TimesNet, and 6.5% for iTransformer.

The 26th European Conference on Artificial Intelligence (ECAI 2023) took place from 30 September – 4 October in Krakow, Poland. On the final day of the conference, the outstanding paper awards were announced. There were two winners in the ECAI 2023 Outstanding Paper category, and one winner in the Outstanding Paper for AI in Social Good category. Abstract: Learning effective strategies in sparse reward tasks is one of the fundamental challenges in reinforcement learning. This becomes extremely difficult in multi-agent environments, as the concurrent learning of multiple agents induces the non-stationarity problem and sharply increased joint state space.

We study the selective learning problem introduced by Qiao and Valiant (2019), in which the learner observes $n$ labeled data points one at a time. At a time of its choosing, the learner selects a window length $w$ and a model $\hat\ell$ from the model class $\mathcal{L}$, and then labels the next $w$ data points using $\hat\ell$. The excess risk incurred by the learner is defined as the difference between the average loss of $\hat\ell$ over those $w$ data points and the smallest possible average loss among all models in $\mathcal{L}$ over those $w$ data points. We give an improved algorithm, termed the hybrid exponential weights algorithm, that achieves an expected excess risk of $O((\log\log|\mathcal{L}| + \log\log n)/\log n)$. This result gives a doubly exponential improvement in the dependence on $|\mathcal{L}|$ over the best known bound of $O(\sqrt{|\mathcal{L}|/\log n})$. We complement the positive result with an almost matching lower bound, which suggests the worst-case optimality of the algorithm. We also study a more restrictive family of learning algorithms that are bounded-recall in the sense that when a prediction window of length $w$ is chosen, the learner's decision only depends on the most recent $w$ data points. We analyze an exponential weights variant of the ERM algorithm in Qiao and Valiant (2019). This new algorithm achieves an expected excess risk of $O(\sqrt{\log |\mathcal{L}|/\log n})$, which is shown to be nearly optimal among all bounded-recall learners. Our analysis builds on a generalized version of the selective mean prediction problem in Drucker (2013); Qiao and Valiant (2019), which may be of independent interest.

Attributes are human describable features, which have been used successfully for face, object, and activity recognition. Facial attributes are intuitive descriptions of faces and have proven to be very useful in face recognition and verification. Despite their usefulness, to date there is only one large-scale facial attribute dataset, CelebA. Impressive results have been achieved on this dataset, but it exhibits a variety of very significant biases. As CelebA contains mostly frontal idealized images of celebrities, it is difficult to generalize a model trained on this data for use on another dataset (of non celebrities). A typical approach to dealing with imbalanced data involves sampling the data in order to balance the positive and negative labels, however, with a multi-label problem this becomes a non-trivial task. By sampling to balance one label, we affect the distribution of other labels in the data. To address this problem, we introduce a novel Selective Learning method for deep networks which adaptively balances the data in each batch according to the desired distribution for each label. The bias in CelebA can be corrected for in this way, allowing the network to learn a more robust attribute model. We argue that without this multi-label balancing, the network cannot learn to accurately predict attributes that are poorly represented in CelebA. We demonstrate the effectiveness of our method on the problem of facial attribute prediction on CelebA, LFWA, and the new University of Maryland Attribute Evaluation Dataset (UMD-AED), outperforming the state-of-the-art on each dataset.

Learning based on networks of real neurons, and by extension biologically inspired models of neural networks, has yet to find general learning rules leading to widespread applications. In this paper, we argue for the existence of a principle allowing to steer the dynamics of a biologically inspired neural network. Using carefully timed external stimulation, the network can be driven towards a desired dynamical state. We term this principle "Learning by Stimulation Avoidance" (LSA). We demonstrate through simulation that the minimal sufficient conditions leading to LSA in artificial networks are also sufficient to reproduce learning results similar to those obtained in biological neurons by Shahaf and Marom [1]. We examine the mechanism's basic dynamics in a reduced network, and demonstrate how it scales up to a network of 100 neurons. We show that LSA has a higher explanatory power than existing hypotheses about the response of biological neural networks to external simulation, and can be used as a learning rule for an embodied application: learning of wall avoidance by a simulated robot. The surge in popularity of artificial neural networks is mostly directed to disembodied models of neurons with biologically irrelevant dynamics: to the authors' knowledge, this is the first work demonstrating sensory-motor learning with random spiking networks through pure Hebbian learning.