Artificial intelligence nowadays plays an increasingly prominent role in our life since decisions that were once made by humans are now delegated to automated systems. A machine learning algorithm trained based on biased data, however, tends to make unfair predictions. Developing classification algorithms that are fair with respect to protected attributes of the data thus becomes an important problem. Motivated by concerns surrounding the fairness effects of sharing and few-shot machine learning tools, such as the Model Agnostic Meta-Learning framework, we propose a novel fair fast-adapted few-shot meta-learning approach that efficiently mitigates biases during meta-train by ensuring controlling the decision boundary covariance that between the protected variable and the signed distance from the feature vectors to the decision boundary. Through extensive experiments on two real-world image benchmarks over three state-of-the-art meta-learning algorithms, we empirically demonstrate that our proposed approach efficiently mitigates biases on model output and generalizes both accuracy and fairness to unseen tasks with a limited amount of training samples.
Deep reinforcement learning has recently seen huge success across multiple areas in the robotics domain. Owing to the limitations of gathering real-world data, i.e., sample inefficiency and the cost of collecting it, simulation environments are utilized for training the different agents. This not only aids in providing a potentially infinite data source, but also alleviates safety concerns with real robots. Nonetheless, the gap between the simulated and real worlds degrades the performance of the policies once the models are transferred into real robots. Multiple research efforts are therefore now being directed towards closing this sim-to-real gap and accomplish more efficient policy transfer. Recent years have seen the emergence of multiple methods applicable to different domains, but there is a lack, to the best of our knowledge, of a comprehensive review summarizing and putting into context the different methods. In this survey paper, we cover the fundamental background behind sim-to-real transfer in deep reinforcement learning and overview the main methods being utilized at the moment: domain randomization, domain adaptation, imitation learning, meta-learning and knowledge distillation. We categorize some of the most relevant recent works, and outline the main application scenarios. Finally, we discuss the main opportunities and challenges of the different approaches and point to the most promising directions.
We uncover an ever-overlooked deficiency in the prevailing Few-Shot Learning (FSL) methods: the pre-trained knowledge is indeed a confounder that limits the performance. This finding is rooted from our causal assumption: a Structural Causal Model (SCM) for the causalities among the pre-trained knowledge, sample features, and labels. Thanks to it, we propose a novel FSL paradigm: Interventional Few-Shot Learning (IFSL). Specifically, we develop three effective IFSL algorithmic implementations based on the backdoor adjustment, which is essentially a causal intervention towards the SCM of many-shot learning: the upper-bound of FSL in a causal view. It is worth noting that the contribution of IFSL is orthogonal to existing fine-tuning and meta-learning based FSL methods, hence IFSL can improve all of them, achieving a new 1-/5-shot state-of-the-art on \textit{mini}ImageNet, \textit{tiered}ImageNet, and cross-domain CUB. Code is released at https://github.com/yue-zhongqi/ifsl.
The problem of learning to generalize to unseen classes during training, known as few-shot classification, has attracted considerable attention. Initialization based methods, such as the gradient-based model agnostic meta-learning (MAML), tackle the few-shot learning problem by "learning to fine-tune". The goal of these approaches is to learn proper model initialization, so that the classifiers for new classes can be learned from a few labeled examples with a small number of gradient update steps. Few shot meta-learning is well-known with its fast-adapted capability and accuracy generalization onto unseen tasks. Learning fairly with unbiased outcomes is another significant hallmark of human intelligence, which is rarely touched in few-shot meta-learning. In this work, we propose a Primal-Dual Fair Meta-learning framework, namely PDFM, which learns to train fair machine learning models using only a few examples based on data from related tasks. The key idea is to learn a good initialization of a fair model's primal and dual parameters so that it can adapt to a new fair learning task via a few gradient update steps. Instead of manually tuning the dual parameters as hyperparameters via a grid search, PDFM optimizes the initialization of the primal and dual parameters jointly for fair meta-learning via a subgradient primal-dual approach. We further instantiate examples of bias controlling using mean difference and decision boundary covariance as fairness constraints to each task for supervised regression and classification, respectively. We demonstrate the versatility of our proposed approach by applying our approach to various real-world datasets. Our experiments show substantial improvements over the best prior work for this setting.
The goal of domain generalization is to learn from multiple source domains to generalize to unseen target domains under distribution discrepancy. Current state-of-the-art methods in this area are fully supervised, but for many real-world problems it is hardly possible to obtain enough labeled samples. In this paper, we propose the first method of domain generalization to leverage unlabeled samples, combining of meta learning's episodic training and semi-supervised learning, called DGSML. DGSML employs an entropy-based pseudo-labeling approach to assign labels to unlabeled samples and then utilizes a novel discrepancy loss to ensure that class centroids before and after labeling unlabeled samples are close to each other. To learn a domain-invariant representation, it also utilizes a novel alignment loss to ensure that the distance between pairs of class centroids, computed after adding the unlabeled samples, is preserved across different domains. DGSML is trained by a meta learning approach to mimic the distribution shift between the input source domains and unseen target domains. Experimental results on benchmark datasets indicate that DGSML outperforms state-of-the-art domain generalization and semi-supervised learning methods.
Faults are endemic to all systems. Adaptive fault-tolerant control maintains degraded performance when faults occur as opposed to unsafe conditions or catastrophic events. In systems with abrupt faults and strict time constraints, it is imperative for control to adapt quickly to system changes to maintain system operations. We present a meta-reinforcement learning approach that quickly adapts its control policy to changing conditions. The approach builds upon model-agnostic meta learning (MAML). The controller maintains a complement of prior policies learned under system faults. This "library" is evaluated on a system after a new fault to initialize the new policy. This contrasts with MAML, where the controller derives intermediate policies anew, sampled from a distribution of similar systems, to initialize a new policy. Our approach improves sample efficiency of the reinforcement learning process. We evaluate our approach on an aircraft fuel transfer system under abrupt faults.
Meta-learning (a.k.a. learning to learn) has recently emerged as a promising paradigm for a variety of applications. There are now many meta-learning methods, each focusing on different modeling aspects of base and meta learners, but all can be (re)formulated as specific bilevel optimization problems. This work presents BOML, a modularized optimization library that unifies several meta-learning algorithms into a common bilevel optimization framework. It provides a hierarchical optimization pipeline together with a variety of iteration modules, which can be used to solve the mainstream categories of meta-learning methods, such as meta-feature-based and meta-initialization-based formulations. The library is written in Python and is available at https://github.com/dut-media-lab/BOML.
3D scene reconstruction from multiple views is an important classical problem in computer vision. Deep learning based approaches have recently demonstrated impressive reconstruction results. When training such models, self-supervised methods are favourable since they do not rely on ground truth data which would be needed for supervised training and is often difficult to obtain. Moreover, learned multi-view stereo reconstruction is prone to environment changes and should robustly generalise to different domains. We propose an adaptive learning approach for multi-view stereo which trains a deep neural network for improved adaptability to new target domains. We use model-agnostic meta-learning (MAML) to train base parameters which, in turn, are adapted for multi-view stereo on new domains through self-supervised training. Our evaluations demonstrate that the proposed adaptation method is effective in learning self-supervised multi-view stereo reconstruction in new domains.
This paper presents a meta-learning based, automatic distribution system load forecasting model selection framework. The framework includes the following processes: feature extraction, candidate model labeling, offline training, and online model recommendation. Using user load forecasting needs as input features, multiple meta-learners are used to rank the available load forecast models based on their forecasting accuracy. Then, a scoring-voting mechanism weights recommendations from each meta-leaner to make the final recommendations. Heterogeneous load forecasting tasks with different temporal and technical requirements at different load aggregation levels are set up to train, validate, and test the performance of the proposed framework. Simulation results demonstrate that the performance of the meta-learning based approach is satisfactory in both seen and unseen forecasting tasks.
We study fairness in supervised few-shot meta-learning models that are sensitive to discrimination (or bias) in historical data. A machine learning model trained based on biased data tends to make unfair predictions for users from minority groups. Although this problem has been studied before, existing methods mainly aim to detect and control the dependency effect of the protected variables (e.g. race, gender) on target prediction based on a large amount of training data. These approaches carry two major drawbacks that (1) lacking showing a global cause-effect visualization for all variables; (2) lacking generalization of both accuracy and fairness to unseen tasks. In this work, we first discover discrimination from data using a causal Bayesian knowledge graph which not only demonstrates the dependency of the protected variable on target but also indicates causal effects between all variables. Next, we develop a novel algorithm based on risk difference in order to quantify the discriminatory influence for each protected variable in the graph. Furthermore, to protect prediction from unfairness, a fast-adapted bias-control approach in meta-learning is proposed, which efficiently mitigates statistical disparity for each task and it thus ensures independence of protected attributes on predictions based on biased and few-shot data samples. Distinct from existing meta-learning models, group unfairness of tasks are efficiently reduced by leveraging the mean difference between (un)protected groups for regression problems. Through extensive experiments on both synthetic and real-world data sets, we demonstrate that our proposed unfairness discovery and prevention approaches efficiently detect discrimination and mitigate biases on model output as well as generalize both accuracy and fairness to unseen tasks with a limited amount of training samples.
The goal of zero-shot learning (ZSL) is to train a model to classify samples of classes that were not seen during training. To address this challenging task, most ZSL methods relate unseen test classes to seen(training) classes via a pre-defined set of attributes that can describe all classes in the same semantic space, so the knowledge learned on the training classes can be adapted to unseen classes. In this paper, we aim to optimize the attribute space for ZSL by training a propagation mechanism to refine the semantic attributes of each class based on its neighbors and related classes on a graph of classes. We show that the propagated attributes can produce classifiers for zero-shot classes with significantly improved performance in different ZSL settings. The graph of classes is usually free or very cheap to acquire such as WordNet or ImageNet classes. When the graph is not provided, given pre-defined semantic embeddings of the classes, we can learn a mechanism to generate the graph in an end-to-end manner along with the propagation mechanism. However, this graph-aided technique has not been well-explored in the literature. In this paper, we introduce the attribute propagation network (APNet), which is composed of 1) a graph propagation model generating attribute vector for each class and 2) a parameterized nearest neighbor (NN) classifier categorizing an image to the class with the nearest attribute vector to the image's embedding. For better generalization over unseen classes, different from previous methods, we adopt a meta-learning strategy to train the propagation mechanism and the similarity metric for the NN classifier on multiple sub-graphs, each associated with a classification task over a subset of training classes. In experiments with two zero-shot learning settings and five benchmark datasets, APNet achieves either compelling performance or new state-of-the-art results.
From CNNs to attention mechanisms, encoding inductive biases into neural networks has been a fruitful source of improvement in machine learning. Adding auxiliary losses to the main objective function is a general way of encoding biases that can help networks learn better representations. However, since auxiliary losses are minimized only on training data, they suffer from the same generalization gap as regular task losses. Moreover, by adding a term to the loss function, the model optimizes a different objective than the one we care about. In this work we address both problems: first, we take inspiration from \textit{transductive learning} and note that after receiving an input but before making a prediction, we can fine-tune our networks on any unsupervised loss. We call this process {\em tailoring}, because we customize the model to each input to ensure our prediction satisfies the inductive bias. Second, we formulate {\em meta-tailoring}, a nested optimization similar to that in meta-learning, and train our models to perform well on the task objective after adapting them using an unsupervised loss. The advantages of tailoring and meta-tailoring are discussed theoretically and demonstrated empirically on a diverse set of examples.
Given an unsupervised outlier detection (OD) task on a new dataset, how can we automatically select a good outlier detection method and its hyperparameter(s) (collectively called a model)? Thus far, model selection for OD has been a "black art"; as any model evaluation is infeasible due to the lack of (i) hold-out data with labels, and (ii) a universal objective function. In this work, we develop the first principled data-driven approach to model selection for OD, called MetaOD, based on meta-learning. MetaOD capitalizes on the past performances of a large body of detection models on existing outlier detection benchmark datasets, and carries over this prior experience to automatically select an effective model to be employed on a new dataset without using any labels. To capture task similarity, we introduce specialized meta-features that quantify outlying characteristics of a dataset. Through comprehensive experiments, we show the effectiveness of MetaOD in selecting a detection model that significantly outperforms the most popular outlier detectors (e.g., LOF and iForest) as well as various state-of-the-art unsupervised meta-learners while being extremely fast. To foster reproducibility and further research on this new problem, we open-source our entire meta-learning system, benchmark environment, and testbed datasets.
Rovers require knowledge of terrain to plan trajectories that maximize safety and efficiency. Terrain type classification relies on input from human operators or machine learning-based image classification algorithms. However, high level terrain classification is typically not sufficient to prevent incidents such as rovers becoming unexpectedly stuck in a sand trap; in these situations, online rover-terrain interaction data can be leveraged to accurately predict future dynamics and prevent further damage to the rover. This paper presents a meta-learning-based approach to adapt probabilistic predictions of rover dynamics by augmenting a nominal model affine in parameters with a Bayesian regression algorithm (P-ALPaCA). A regularization scheme is introduced to encourage orthogonality of nominal and learned features, leading to interpretable probabilistic estimates of terrain parameters in varying terrain conditions.
Decisions made by machine learning systems have increasing influence on the world, yet it is common for machine learning algorithms to assume that no such influence exists. An example is the use of the i.i.d. assumption in content recommendation. In fact, the (choice of) content displayed can change users' perceptions and preferences, or even drive them away, causing a shift in the distribution of users. We introduce the term auto-induced distributional shift (ADS) to describe the phenomenon of an algorithm causing a change in the distribution of its own inputs. Our goal is to ensure that machine learning systems do not leverage ADS to increase performance when doing so could be undesirable. We demonstrate that changes to the learning algorithm, such as the introduction of meta-learning, can cause hidden incentives for auto-induced distributional shift (HI-ADS) to be revealed. To address this issue, we introduce `unit tests' and a mitigation strategy for HI-ADS, as well as a toy environment for modelling real-world issues with HI-ADS in content recommendation, where we demonstrate that strong meta-learners achieve gains in performance via ADS. We show meta-learning and Q-learning both sometimes fail unit tests, but pass when using our mitigation strategy.
One-hot labels do not represent soft decision boundaries among concepts, and hence, models trained on them are prone to overfitting. Using soft labels as targets provide regularization, but different soft labels might be optimal at different stages of optimization. Also, training with fixed labels in the presence of noisy annotations leads to worse generalization. To address these limitations, we propose a framework, where we treat the labels as learnable parameters, and optimize them along with model parameters. The learned labels continuously adapt themselves to the model's state, thereby providing dynamic regularization. When applied to the task of supervised image-classification, our method leads to consistent gains across different datasets and architectures. For instance, dynamically learned labels improve ResNet18 by 2.1% on CIFAR100. When applied to dataset containing noisy labels, the learned labels correct the annotation mistakes, and improves over state-of-the-art by a significant margin. Finally, we show that learned labels capture semantic relationship between classes, and thereby improve teacher models for the downstream task of distillation.
Training deep learning models on embedded devices is typically avoided since this requires more memory, computation and power over inference. In this work, we focus on lowering the amount of memory needed for storing all activations, which are required during the backward pass to compute the gradients. Instead, during the forward pass, static Synthetic Gradient Modules (SGMs) predict gradients for each layer. This allows training the model in a feed-forward manner without having to store all activations. We tested our method on a robot grasping scenario where a robot needs to learn to grasp new objects given only a single demonstration. By first training the SGMs in a meta-learning manner on a set of common objects, during fine-tuning, the SGMs provided the model with accurate gradients to successfully learn to grasp new objects. We have shown that our method has comparable results to using standard backpropagation.
The recent outbreak of COVID-19 has affected millions of individuals around the world and has posed a significant challenge to global healthcare. From the early days of the pandemic, it became clear that it is highly contagious and that human mobility contributes significantly to its spread. In this paper, we study the impact of population movement on the spread of COVID-19, and we capitalize on recent advances in the field of representation learning on graphs to capture the underlying dynamics. Specifically, we create a graph where nodes correspond to a country's regions and the edge weights denote human mobility from one region to another. Then, we employ graph neural networks to predict the number of future cases, encoding the underlying diffusion patterns that govern the spread into our learning model. Furthermore, to account for the limited amount of training data, we capitalize on the pandemic's asynchronous outbreaks across countries and use a model-agnostic meta-learning based method to transfer knowledge from one country's model to another's. We compare the proposed approach against simple baselines and more traditional forecasting techniques in 3 European countries. Experimental results demonstrate the superiority of our method, highlighting the usefulness of GNNs in epidemiological prediction. Transfer learning provides the best model, highlighting its potential to improve the accuracy of the predictions in case of secondary waves, if data from past/parallel outbreaks is utilized.
As data from IoT (Internet of Things) sensors become ubiquitous, state-of-the-art machine learning algorithms face many challenges on directly using sensor data. To overcome these challenges, methods must be designed to learn directly from sensors without manual annotations. This paper introduces Sensory Time-cue for Unsupervised Meta-learning (STUM). Different from traditional learning approaches that either heavily depend on labels or on time-independent feature extraction assumptions, such as Gaussian distribution features, the STUM system uses time relation of inputs to guide the feature space formation within and across modalities. The fact that STUM learns from a variety of small tasks may put this method in the camp of Meta-Learning. Different from existing Meta-Learning approaches, STUM learning tasks are composed within and across multiple modalities based on time-cue co-exist with the IoT streaming data. In an audiovisual learning example, because consecutive visual frames usually comprise the same object, this approach provides a unique way to organize features from the same object together. The same method can also organize visual object features with the object's spoken-name features together if the spoken name is presented with the object at about the same time. This cross-modality feature organization may further help the organization of visual features that belong to similar objects but acquired at different location and time. Promising results are achieved through evaluations.
In real-world applications, data do not reflect the ones commonly used for neural networks training, since they are usually few, unlabeled and can be available as a stream. Hence many existing deep learning solutions suffer from a limited range of applications, in particular in the case of online streaming data that evolve over time. To narrow this gap, in this work we introduce a novel and complex setting involving unsupervised meta-continual learning with unbalanced tasks. These tasks are built through a clustering procedure applied to a fitted embedding space. We exploit a meta-learning scheme that simultaneously alleviates catastrophic forgetting and favors the generalization to new tasks. Moreover, to encourage feature reuse during the meta-optimization, we exploit a single inner loop taking advantage of an aggregated representation achieved through the use of a self-attention mechanism. Experimental results on few-shot learning benchmarks show competitive performance even compared to the supervised case. Additionally, we empirically observe that in an unsupervised scenario, the small tasks and the variability in the clusters pooling play a crucial role in the generalization capability of the network. Further, on complex datasets, the exploitation of more clusters than the true number of classes leads to higher results, even compared to the ones obtained with full supervision, suggesting that a predefined partitioning into classes can miss relevant structural information.