This paper proposes a cooperative learning algorithm to train both the undirected energy-based model and the directed latent variable model jointly. The learning algorithm interweaves the maximum likelihood algorithms for learning the two models, and each iteration consists of the following two steps: (1) Modified contrastive divergence for energy-based model: The learning of the energy-based model is based on the contrastive divergence, but the finite-step MCMC sampling of the model is initialized from the synthesized examples generated by the latent variable model instead of being initialized from the observed examples. (2) MCMC teaching of the latent variable model: The learning of the latent variable model is based on how the MCMC in (1) changes the initial synthesized examples generated by the latent variable model, where the latent variables that generate the initial synthesized examples are known so that the learning is essentially supervised. Our experiments show that the cooperative learning algorithm can learn realistic models of images.

Effective methods for learning deep neural networks with fewer parameters are urgently required, since storage and computations of heavy neural networks have largely prevented their widespread use on mobile devices. Compared with algorithms which directly remove weights or filters for obtaining considerable compression and speed-up ratios, training thin deep networks exploiting the student-teacher learning paradigm is more flexible. However, it is very hard to determine which formulation is optimal to measure the information inherited from teacher networks. To overcome this challenge, we utilize the generative adversarial network (GAN) to learn the student network. In practice, the generator is exactly the student network with extremely less parameters and the discriminator is used as a teaching assistant for distinguishing features extracted from student and teacher networks. By simultaneously optimizing the generator and the discriminator, the resulting student network can produce features of input data with the similar distribution as that of features of the teacher network. Extensive experimental results on benchmark datasets demonstrate that the proposed method is capable of learning well-performed portable networks, which is superior to the state-of-the-art methods.

Lifelong learning intends to learn new consecutive tasks depending on previously accumulated experiences, i.e., knowledge library. However, the knowledge among different new coming tasks are imbalance. Therefore, in this paper, we try to mimic an effective "human cognition" strategy by actively sorting the importance of new tasks in the process of unknown-to-known and selecting to learn the important tasks with more information preferentially. To achieve this, we consider to assess the importance of the new coming task, i.e., unknown or not, as an outlier detection issue, and design a hierarchical dictionary learning model consisting of two-level task descriptors to sparse reconstruct each task with the l0 norm constraint. The new coming tasks are sorted depending on the sparse reconstruction score in descending order, and the task with high reconstruction score will be permitted to pass, where this mechanism is called as "watchdog." Next, the knowledge library of the lifelong learning framework encode the selected task by transferring previous knowledge, and then can also update itself with knowledge from both previously learned task and current task automatically. For model optimization, the alternating direction method is employed to solve our model and converges to a fixed point. Extensive experiments on both benchmark datasets and our own dataset demonstrate the effectiveness of our proposed model especially in task selection and dictionary learning.

Bayesian optimisation has been successfully applied to a variety of reinforcement learning problems. However, the traditional approach for learning optimal policies in simulators does not utilise the opportunity to improve learning by adjusting certain environment variables: state features that are unobservable and randomly determined by the environment in a physical setting but are controllable in a simulator. This paper considers the problem of finding a robust policy while taking into account the impact of environment variables. We present Alternating Optimisation and Quadrature (ALOQ), which uses Bayesian optimisation and Bayesian quadrature to address such settings. ALOQ is robust to the presence of significant rare events, which may not be observable under random sampling, but play a substantial role in determining the optimal policy. Experimental results across different domains show that ALOQ can learn more efficiently and robustly than existing methods.

We consider a new setting of online clustering of contextual cascading bandits, an online learning problem where the underlying cluster structure over users is unknown and needs to be learned from a random prefix feedback. More precisely, a learning agent recommends an ordered list of items to a user, who checks the list and stops at the first satisfactory item, if any. We propose an algorithm of CLUB-cascade for this setting and prove an n-step regret bound of order O(√n). Previous work corresponds to the degenerate case of only one cluster, and our general regret bound in this special case also significantly improves theirs. We conduct experiments on both synthetic and real data, and demonstrate the effectiveness of our algorithm and the advantage of incorporating online clustering method.

Mixed data with both categorical and continuous features are ubiquitous in real-world applications. Learning a good representation of mixed data is critical yet challenging for further learning tasks. Existing methods for representing mixed data often overlook the heterogeneous coupling relationships between categorical and continuous features as well as the discrimination between objects. To address these issues, we propose an auto-instructive representation learning scheme to enable margin-enhanced distance metric learning for a discrimination-enhanced representation. Accordingly, we design a metric-based auto-instructor (MAI) model which consists of two collaborative instructors. Each instructor captures the feature-level couplings in mixed data with fully connected networks, and guides the infinite-margin metric learning for the peer instructor with a contrastive order. By feeding the learned representation into both partition-based and density-based clustering methods, our experiments on eight UCI datasets show highly significant learning performance improvement and much more distinguishable visualization outcomes over the baseline methods.

Deep reinforcement learning has emerged as a powerful tool for a variety of learning tasks, however deep nets typically exhibit forgetting when learning multiple tasks in sequence. To mitigate forgetting, we propose an experience replay process that augments the standard FIFO buffer and selectively stores experiences in a long-term memory. We explore four strategies for selecting which experiences will be stored: favoring surprise, favoring reward, matching the global training distribution, and maximizing coverage of the state space. We show that distribution matching successfully prevents catastrophic forgetting, and is consistently the best approach on all domains tested. While distribution matching has better and more consistent performance, we identify one case in which coverage maximization is beneficial---when tasks that receive less trained are more important. Overall, our results show that selective experience replay, when suitable selection algorithms are employed, can prevent catastrophic forgetting.

Deep reinforcement learning (RL) has achieved several high profile successes in difficult decision-making problems. However, these algorithms typically require a huge amount of data before they reach reasonable performance. In fact, their performance during learning can be extremely poor. This may be acceptable for a simulator, but it severely limits the applicability of deep RL to many real-world tasks, where the agent must learn in the real environment. In this paper we study a setting where the agent may access data from previous control of the system. We present an algorithm, Deep Q-learning from Demonstrations (DQfD), that leverages small sets of demonstration data to massively accelerate the learning process even from relatively small amounts of demonstration data and is able to automatically assess the necessary ratio of demonstration data while learning thanks to a prioritized replay mechanism. DQfD works by combining temporal difference updates with supervised classification of the demonstrator’s actions. We show that DQfD has better initial performance than Prioritized Dueling Double Deep Q-Networks (PDD DQN) as it starts with better scores on the first million steps on 41 of 42 games and on average it takes PDD DQN 83 million steps to catch up to DQfD’s performance. DQfD learns to out-perform the best demonstration given in 14 of 42 games. In addition, DQfD leverages human demonstrations to achieve state-of-the-art results for 11 games. Finally, we show that DQfD performs better than three related algorithms for incorporating demonstration data into DQN.

Linear regression methods are commonly used by both researchers and data scientists due to their interpretability and their reduced likelihood of overfitting. However, these methods can still perform poorly if little labeled training data is available. Typical methods used to overcome a lack of labeled training data somehow involve exploiting an outside source of labeled data or large amounts of unlabeled data. This includes areas such as active learning, semi-supervised learning and transfer learning, but in many domains these approaches are not always applicable because they require either a mechanism to label data, large amounts of unlabeled data or additional sources of sufficiently related data. In this paper we explore an alternative, non-data centric approach. We allow the user to guide the learning system through three forms of feature-level guidance which constrain the parameters of the regression function. Such guidance is unlikely to be perfectly accurate, so we derive methods which are robust to some amounts of noise, a property we formally prove for one of our methods.

We tackle the problem of constructive preference elicitation, that is the problem of learning user preferences over verylarge decision problems, involving a combinatorial space of possible outcomes. In this setting, the suggested configuration is synthesized on-the-fly by solving a constrained optimization problem, while the preferences are learned iteratively by interacting with the user. Previous work has shown that Coactive Learning is a suitable method for learning userpreferences in constructive scenarios. In Coactive Learning the user provides feedback to the algorithm in the form of an improvement to a suggested configuration. When the problem involves many decision variables and constraints, this type of interaction poses a significant cognitive burden on the user. We propose a decomposition technique for large preference-based decision problems relying exclusively on inference and feedback over partial configurations. This has the clear advantage of drastically reducing the user cognitive load. Additionally, part-wise inference can be (up to exponentially) less computationally demanding than inference over full configurations. We discuss the theoretical implications of working with parts and present promising empirical results on one synthetic and two realistic constructive problems.