1,721,049 research outputs found
Graph Neural Network for Context-Aware Recommendation
No full textRecommendation problems are naturally tackled as a link prediction task in a bipartite graph between user and item nodes, labelled with rating information on edges. To provide personal recommendations and improve the performance of the recommender system, it is necessary to integrate side information along with user-item interactions. The integration of context is a key success factor in recommendation systems because it allows catering for user preferences and opinions, especially when this pertains to the circumstances surrounding the interaction between users and items. In this paper, we propose a context-aware Graph Convolutional Matrix Completion which captures structural information and integrates the user's opinion on items along with the surrounding context on edges and static features of user and item nodes. Our graph encoder produces user and item representations with respect to context, features and opinion. The decoder takes the aggregated embeddings to predict the user-item score considering the surrounding context. We have evaluated the performance of our model on 14 five publicly available datasets and compared it with state-of-the-art algorithms. Throughout this we show how it can effectively integrate user opinion along with surrounding context to produce a final node representation which is aware of the favourite circumstances of the particular node
A tensor framework for learning in structured domains
No full textLearning machines for structured data (e.g., trees) are intrinsically based on their capacity to learn representations by aggregating information from the multi-way relationships emerging from the structure topology. While complex aggregation functions are desirable in this context to increase the expressiveness of the learned representations, the modelling of higher-order interactions among structure constituents is unfeasible, in practice, due to the exponential number of parameters required. Therefore, the common approach is to define models which rely only on first-order interactions among structure constituents. In this work, we leverage tensors theory to define a framework for learning in structured domains. Such a framework is built on the observation that more expressive models require a tensor parameterisation. This observation is the stepping stone for the application of tensor decompositions in the context of recursive models. From this point of view, the advantage of using tensor decompositions is twofold since it allows limiting the number of model parameters while injecting inductive biases that do not ignore higher-order interactions. We apply the proposed framework on probabilistic and neural models for structured data, defining different models which leverage tensor decompositions. The experimental validation clearly shows the advantage of these models compared to first-order and full-tensorial models
Bayesian Tensor Factorisation for Bottom-up Hidden Tree Markov Models
No full textBottom-Up Hidden Tree Markov Model is a highly expressive model for tree-structured data. Unfortunately, it cannot be used in practice due to the intractable size of its state-transition matrix. We propose a new approximation which lies on the Tucker factorisation of tensors. The probabilistic interpretation of such approximation allows us to define a new probabilistic model for tree-structured data. Hence, we define the new approximated model and we derive its learning algorithm. Then, we empirically assess the effective power of the new model evaluating it on two different tasks. In both cases, our model outperforms the other approximated model known in the literature
Context-Aware Graph Convolutional Autoencoder
No full textRecommendation problems can be addressed as link prediction tasks in a bipartite graph between user and item nodes, labelled with rating on edges. Existing matrix completion approaches model the user’s opinion on items by ignoring context information that can instead be associated with the edges of the bipartite graph. Context is an important factor to be considered as it heavily affects opinions and preferences. Following this line of research, this paper proposes a graph convolutional auto-encoder approach which considers users’ opinion on items as well as the static node features and context information on edges. Our graph encoder produces a representation of users and items from the perspective of context, static features, and rating opinion. The empirical analysis on three real-world datasets shows that the proposed approach outperforms recent state-of-the-art recommendation systems
Leveraging Relational Information for Learning Weakly Disentangled Representations
No full textDisentanglement is a difficult property to enforce in neural representations. This might be due, in part, to a formalization of the disentanglement problem that focuses too heavily on separating relevant factors of variation of the data in single isolated dimensions of the neural representation. We argue that such a definition might be too restrictive and not necessarily beneficial in terms of downstream tasks. In this work, we present an alternative view over learning (weakly) disentangled representations, which leverages concepts from relational learning. We identify the regions of the latent space that correspond to specific instances of generative factors, and we learn the relationships among these regions in order to perform controlled changes to the latent codes. We also introduce a compound generative model that implements such a weak disentanglement approach. Our experiments shows that the learned representations can separate the relevant factors of variation in the data, while preserving the information needed for effectively generating high quality data samples
Tensor decompositions in recursive neural networks for tree-structured data
No full textThe paper introduces two new aggregation functions to encode structural knowledge from tree-structured data. They leverage the Canonical and Tensor-Train decompositions to yield expressive context aggregation while limiting the number of model parameters. Finally, we define two novel neural recursive models for trees leveraging such aggregation functions, and we test them on two tree classification tasks, showing the advantage of proposed models when tree outdegree increases
Generalising Recursive Neural Models by Tensor Decomposition
No full textMost machine learning models for structured data encode the structural knowledge of a node by leveraging simple aggregation functions (in neural models, typically a weighted sum) of the information in the node's neighbourhood. Nevertheless, the choice of simple context aggregation functions, such as the sum, can be widely sub-optimal. In this work we introduce a general approach to model aggregation of structural context leveraging a tensor-based formulation. We show how the exponential growth in the size of the parameter space can be controlled through an approximation based on the Tucker tensor decomposition. This approximation allows limiting the parameters space size, decoupling it from its strict relation with the size of the hidden encoding space. By this means, we can effectively regulate the trade-off between expressivity of the encoding, controlled by the hidden size, computational complexity and model generalisation, influenced by parameterisation. Finally, we introduce a new Tensorial Tree-LSTM derived as an instance of our framework and we use it to experimentally assess our working hypotheses on tree classification scenarios
Tensor decompositions in deep learning
No full textThe paper surveys the topic of tensor decompositions in modern machine learning applications. It focuses on three active research topics of significant relevance for the community. After a brief review of consolidated works on multi-way data analysis, we consider the use of tensor decompositions in compressing the parameter space of deep learning models. Lastly, we discuss how tensor methods can be leveraged to yield richer adaptive representations of complex data, including structured information. The paper concludes with a discussion on interesting open research challenges
Continual adaptation of federated reservoirs in pervasive environments
No full textWhen performing learning tasks in pervasive environments, the main challenge arises from the need of combining federated and continual settings. The former comes from the massive distribution of devices with privacy-regulated data. The latter is required by the low resources of the participating devices, which may retain data for short periods of time. In this paper, we propose a setup for learning with Echo State Networks (ESNs) in pervasive environments. Our proposal focuses on the use of Intrinsic Plasticity (IP), a gradient-based method for adapting the reservoir's non-linearity. First, we extend the objective function of IP to include the uncertainty arising from the distribution of the data over space and time. Then, we propose Federated Intrinsic Plasticity (FedIP), which is intended for client–server federated topologies with stationary data, and adapts the learning scheme provided by Federated Averaging (FedAvg) to include the learning rule of IP. Finally, we further extend this algorithm for learning to Federated Continual Intrinsic Plasticity (FedCLIP) to equip clients with CL strategies for dealing with continuous data streams. We evaluate our approach on an incremental setup built upon real-world datasets from human monitoring, where we tune the complexity of the scenario in terms of the distribution of the data over space and time. Results show that both our algorithms improve the representation capabilities and the performance of the ESN, while being robust to catastrophic forgetting
Encoding-based memory for recurrent neural networks
No full textLearning to solve sequential tasks with recurrent models requires the ability to memorize long sequences and to extract task-relevant features from them. In this paper, we study memorization from the point of view of the design and training of recurrent neural networks. We study how to maximize the short-term memory of recurrent units, an objective difficult to achieve using backpropagation. We propose a new model, the Linear Memory Network, which features an encoding-based memorization component built with a linear autoencoder for sequences. Additionally, we provide a specialized training algorithm that initializes the memory to efficiently encode the hidden activations of the network. Experimental results on synthetic and real-world datasets show that the chosen encoding mechanism is superior to static encodings such as orthogonal models and the delay line. The method also outperforms RNN and LSTM units trained using stochastic gradient descent. Experiments on symbolic music modeling show that the training algorithm specialized for the memorization component improves the final performance compared to stochastic gradient descent
- …
