Machine learning can be applied to time series datasets. These are problems where a numeric or categorical value must be predicted, but the rows of data are ordered by time. A problem when getting started in time series forecasting with machine learning is finding good quality standard datasets on which to practice. In this post, […]

Driven by the increasingly larger deep learning models, neural language generation (NLG) has enjoyed unprecedentedly improvement and is now able to generate a diversity of human-like texts on demand, granting itself the capability of serving as a human writing assistant. Text attribute transfer is one of the most important NLG tasks, which aims to control certain attributes that people may expect the texts to possess, such as sentiment, tense, emotion, political position, etc. It has a long history in Natural Language Processing but recently gains much more attention thanks to the promising performance brought by deep learning models. In this article, we present a systematic survey on these works for neural text attribute transfer. We collect all related academic works since the first appearance in 2017. We then select, summarize, discuss, and analyze around 65 representative works in a comprehensive way. Overall, we have covered the task formulation, existing datasets and metrics for model development and evaluation, and all methods developed over the last several years. We reveal that existing methods are indeed based on a combination of several loss functions with each of which serving a certain goal. Such a unique perspective we provide could shed light on the design of new methods. We conclude our survey with a discussion on open issues that need to be resolved for better future development.

The classification of DNA sequences is a key research area in bioinformatics as it enables researchers to conduct genomic analysis and detect possible diseases. In this paper, three state-of-the-art algorithms, namely Convolutional Neural Networks, Deep Neural Networks, and N-gram Probabilistic Models, are used for the task of DNA classification. Furthermore, we introduce a novel feature extraction method based on the Levenshtein distance and randomly generated DNA sub-sequences to compute information-rich features from the DNA sequences. We also use an existing feature extraction method based on 3-grams to represent amino acids and combine both feature extraction methods with a multitude of machine learning algorithms. Four different data sets, each concerning viral diseases such as Covid-19, AIDS, Influenza, and Hepatitis C, are used for evaluating the different approaches. The results of the experiments show that all methods obtain high accuracies on the different DNA datasets. Furthermore, the domain-specific 3-gram feature extraction method leads in general to the best results in the experiments, while the newly proposed technique outperforms all other methods on the smallest Covid-19 dataset

The stochastic block model is one of the most studied network models for community detection. It is well-known that most algorithms proposed for fitting the stochastic block model likelihood function cannot scale to large-scale networks. One prominent work that overcomes this computational challenge is Amini et al.(2013), which proposed a fast pseudo-likelihood approach for fitting stochastic block models to large sparse networks. However, this approach does not have convergence guarantee, and is not well suited for small- or medium- scale networks. In this article, we propose a novel likelihood based approach that decouples row and column labels in the likelihood function, which enables a fast alternating maximization; the new method is computationally efficient, performs well for both small and large scale networks, and has provable convergence guarantee. We show that our method provides strongly consistent estimates of the communities in a stochastic block model. As demonstrated in simulation studies, the proposed method outperforms the pseudo-likelihood approach in terms of both estimation accuracy and computation efficiency, especially for large sparse networks. We further consider extensions of our proposed method to handle networks with degree heterogeneity and bipartite properties.

Ordinary (pointwise) binary classification aims to learn a binary classifier from pointwise labeled data. However, such pointwise labels may not be directly accessible due to privacy, confidentiality, or security considerations. In this case, can we still learn an accurate binary classifier? This paper proposes a novel setting, namely pairwise comparison (Pcomp) classification, where we are given only pairs of unlabeled data that we know one is more likely to be positive than the other, instead of pointwise labeled data. Pcomp classification is useful for private or subjective classification tasks. To solve this problem, we present a mathematical formulation for the generation process of pairwise comparison data, based on which we exploit an unbiased risk estimator (URE) to train a binary classifier by empirical risk minimization and establish an estimation error bound. We first prove that a URE can be derived and improve it using correction functions. Then, we start from the noisy-label learning perspective to introduce a progressive URE and improve it by imposing consistency regularization. Finally, experiments validate the effectiveness of our proposed solutions for Pcomp classification.

The neural attention mechanism plays an important role in many natural language processing applications. In particular, the use of multi-head attention extends single-head attention by allowing a model to jointly attend information from different perspectives. Without explicit constraining, however, multi-head attention may suffer from attention collapse, an issue that makes different heads extract similar attentive features, thus limiting the model's representation power. In this paper, for the first time, we provide a novel understanding of multi-head attention from a Bayesian perspective. Based on the recently developed particle-optimization sampling techniques, we propose a non-parametric approach that explicitly improves the repulsiveness in multi-head attention and consequently strengthens model's expressiveness. Remarkably, our Bayesian interpretation provides theoretical inspirations on the not-well-understood questions: why and how one uses multi-head attention. Extensive experiments on various attention models and applications demonstrate that the proposed repulsive attention can improve the learned feature diversity, leading to more informative representations with consistent performance improvement on various tasks.

Bayesian optimisation is a popular method for efficient optimisation of expensive black-box functions. Traditionally, BO assumes that the search space is known. However, in many problems, this assumption does not hold. To this end, we propose a novel BO algorithm which expands (and shifts) the search space over iterations based on controlling the expansion rate thought a hyperharmonic series. Further, we propose another variant of our algorithm that scales to high dimensions. We show theoretically that for both our algorithms, the cumulative regret grows at sub-linear rates. Our experiments with synthetic and real-world optimisation tasks demonstrate the superiority of our algorithms over the current state-of-the-art methods for Bayesian optimisation in unknown search space.

Social media has become an important tool to share information about crisis events such as natural disasters and mass attacks. Detecting actionable posts that contain useful information requires rapid analysis of huge volume of data in real-time. This poses a complex problem due to the large amount of posts that do not contain any actionable information. Furthermore, the classification of information in real-time systems requires training on out-of-domain data, as we do not have any data from a new emerging crisis. Prior work focuses on models pre-trained on similar event types. However, those models capture unnecessary event-specific biases, like the location of the event, which affect the generalizability and performance of the classifiers on new unseen data from an emerging new event. In our work, we train an adversarial neural model to remove latent event-specific biases and improve the performance on tweet importance classification.

Legged robots often use separate control policies that are highly engineered for traversing difficult terrain such as stairs, gaps, and steps, where switching between policies is only possible when the robot is in a region that is common to adjacent controllers. Deep Reinforcement Learning (DRL) is a promising alternative to hand-crafted control design, though typically requires the full set of test conditions to be known before training. DRL policies can result in complex (often unrealistic) behaviours that have few or no overlapping regions between adjacent policies, making it difficult to switch behaviours. In this work we develop multiple DRL policies with Curriculum Learning (CL), each that can traverse a single respective terrain condition, while ensuring an overlap between policies. We then train a network for each destination policy that estimates the likelihood of successfully switching from any other policy. We evaluate our switching method on a previously unseen combination of terrain artifacts and show that it performs better than heuristic methods. While our method is trained on individual terrain types, it performs comparably to a Deep Q Network trained on the full set of terrain conditions. This approach allows the development of separate policies in constrained conditions with embedded prior knowledge about each behaviour, that is scalable to any number of behaviours, and prepares DRL methods for applications in the real world

NN variants have been designed to increase performance in certain problem domains; the convolutional neural network (CNN) excels in the context of image-based tasks, and the recurrent neural network (RNN) in the space of natural language processing and time series analysis. NNs have also been leveraged as components in composite DL frameworks -- they have been used as trainable generators and discriminators in generative adversarial networks (GANs), and as encoders and decoders in transformers [46]. Although they seem unrelated, the images used as inputs in computer vision, and the sentences used as inputs in natural language processing can both be represented by a single, general data structure: the graph (see Figure 1). Formally, a graph is a set of distinct vertices (representing items or entities) that are joined optionally to each other by edges (representing relationships). The learning architecture that has been designed to process said graphs is the titular graph neural network (GNN). Uniquely, the graphs fed into a GNN (during training and evaluation) do not have strict structural requirements per se; the number of vertices and edges between input graphs can change. In this way, GNNs can handle unstructured, non-Euclidean data [4], a property which makes them valuable in certain problem domains where graph data is abundant. Conversely, NN-based algorithms are typically required to operate on structured inputs with strictly defined dimensions.