Recent studies have shown that the choice of activation function can significantly affect the performance of deep learning networks. However, the benefits of novel activation functions have been inconsistent and task dependent, and therefore the rectified linear unit (ReLU) is still the most commonly used. This paper proposes a technique for customizing activation functions automatically, resulting in reliable improvements in performance. Evolutionary search is used to discover the general form of the function, and gradient descent to optimize its parameters for different parts of the network and over the learning process. Experiments with four different neural network architectures on the CIFAR-10 and CIFAR-100 image classification datasets show that this approach is effective. It discovers both general activation functions and specialized functions for different architectures, consistently improving accuracy over ReLU and other recently proposed activation functions by significant margins. The approach can therefore be used as an automated optimization step in applying deep learning to new tasks. The rectified linear unit (ReLU(x) max{x, 0}) is the most commonly used activation function in modern deep learning architectures (Nair & Hinton, 2010). When introduced, it offered substantial improvements over the previously popular tanh and sigmoid activation functions. Because ReLU is unbounded as x, it is less susceptible to vanishing gradients than tanh and sigmoid are. It is also simple to calculate, which leads to faster training times. Activation function design continues to be an active area of research, and a number of novel activation functions have been introduced since ReLU, each with different properties (Nwankpa et al., 2018).

The celebrated Seq2Seq technique and its numerous variants achieve excellent performance on many tasks such as neural machine translation, semantic parsing, and math word problem solving. However, these models either only consider input objects as sequences while ignoring the important structural information for encoding, or they simply treat output objects as sequence outputs instead of structural objects for decoding. In this paper, we present a novel Graph-to-Tree Neural Networks, namely Graph2Tree consisting of a graph encoder and a hierarchical tree decoder, that encodes an augmented graph-structured input and decodes a tree-structured output. In particular, we investigated our model for solving two problems, neural semantic parsing and math word problem. Our extensive experiments demonstrate that our Graph2Tree model outperforms or matches the performance of other state-of-the-art models on these tasks.

Topic segmentation is critical in key NLP tasks and recent works favor highly effective neural supervised approaches. However, current neural solutions are arguably limited in how they model context. In this paper, we enhance a segmenter based on a hierarchical attention BiLSTM network to better model context, by adding a coherence-related auxiliary task and restricted self-attention. Our optimized segmenter outperforms SOTA approaches when trained and tested on three datasets. We also the robustness of our proposed model in domain transfer setting by training a model on a large-scale dataset and testing it on four challenging real-world benchmarks. Furthermore, we apply our proposed strategy to two other languages (German and Chinese), and show its effectiveness in multilingual scenarios.

Automatic charge prediction task aims to determine the final charges based on fact descriptions of criminal cases, which is a vital application of legal assistant systems. Conventional works usually depend on fact descriptions to predict charges while ignoring the legal schematic knowledge, which makes it difficult to distinguish confusing charges. In this paper, we propose a knowledge-attentive neural network model, which introduces legal schematic knowledge about charges and exploit the knowledge hierarchical representation as the discriminative features to differentiate confusing charges. Our model takes the textual fact description as the input and learns fact representation through a graph convolutional network. A legal schematic knowledge transformer is utilized to generate crucial knowledge representations oriented to the legal schematic knowledge at both the schema and charge levels. We apply a knowledge matching network for effectively incorporating charge information into the fact to learn knowledge-aware fact representation. Finally, we use the knowledge-aware fact representation for charge prediction. We create two real-world datasets and experimental results show that our proposed model can outperform other state-of-the-art baselines on accuracy and F1 score, especially on dealing with confusing charges.

Humans can learn and reason under substantial uncertainty in a space of infinitely many concepts, including structured relational concepts ("a scene with objects that have the same color") and ad-hoc categories defined through goals ("objects that could fall on one's head"). In contrast, standard classification benchmarks: 1) consider only a fixed set of category labels, 2) do not evaluate compositional concept learning and 3) do not explicitly capture a notion of reasoning under uncertainty. We introduce a new few-shot, meta-learning benchmark, Compositional Reasoning Under Uncertainty (CURI) to bridge this gap. CURI evaluates different aspects of productive and systematic generalization, including abstract understandings of disentangling, productive generalization, learning boolean operations, variable binding, etc. Importantly, it also defines a model-independent "compositionality gap" to evaluate the difficulty of generalizing out-of-distribution along each of these axes. Extensive evaluations across a range of modeling choices spanning different modalities (image, schemas, and sounds), splits, privileged auxiliary concept information, and choices of negatives reveal substantial scope for modeling advances on the proposed task. All code and datasets will be available online.

Recent work by Clark et al. (2020) shows that transformers can act as 'soft theorem provers' by answering questions over explicitly provided knowledge in natural language. In our work, we take a step closer to emulating formal theorem provers, by proposing PROVER, an interpretable transformer-based model that jointly answers binary questions over rule-bases and generates the corresponding proofs. Our model learns to predict nodes and edges corresponding to proof graphs in an efficient constrained training paradigm. During inference, a valid proof, satisfying a set of global constraints is generated. We conduct experiments on synthetic, hand-authored, and human-paraphrased rule-bases to show promising results for QA and proof generation, with strong generalization performance. First, PROVER generates proofs with an accuracy of 87%, while retaining or improving performance on the QA task, compared to RuleTakers (up to 6% improvement on zero-shot evaluation). Second, when trained on questions requiring lower depths of reasoning, it generalizes significantly better to higher depths (up to 15% improvement). Third, PROVER obtains near perfect QA accuracy of 98% using only 40% of the training data. However, generating proofs for questions requiring higher depths of reasoning becomes challenging, and the accuracy drops to 65% for 'depth 5', indicating significant scope for future work. Our code and models are publicly available at https://github.com/swarnaHub/PRover

Sadly, no other animal represents the world with language or drawing. Early examples of drawing date back to 60,000 years ago [19], though red pigments for mark making are already found 200,000 years ago in the middle stone age [18]. "The first man to make a mammoth appear on the wall of a cave was, I am confident, amazed by what he had done" writes Gibson, because they had discovered that by means of lines they could delineate something [10] p263. What allowed humans to learn to create (visual) abstractions, e.g., the Western child's human stick figure, the Australian aboriginal top-down projections of people seated around a fireplace, the Egyptian formalism for representing things in orthographic projection with multiple station-points, and with social dominance relations transformed into size differences, or the 16th Century Japanese affine projections that have a birdseye viewpoint? Making our own abstraction creating (drawing) machines is one way to find out the answer. A wonderful start was made by Harold Cohen's abstract drawing programs "Aaron" [6]. Aaron and Harold produced beautiful and interesting abstracted drawings that looked as if they had been made by a human alone. But it had no learning, was not conditioned on looking at the world, and was an entirely hand designed production system (a complex set of hierarchical rules for drawing).

Sequence classification is the task of predicting a class label given a sequence of observations. In many applications such as healthcare monitoring or intrusion detection, early classification is crucial to prompt intervention. In this work, we learn sequence classifiers that favour early classification from an evolving observation trace. While many state-of-the-art sequence classifiers are neural networks, and in particular LSTMs, our classifiers take the form of finite state automata and are learned via discrete optimization. Our automata-based classifiers are interpretable---supporting explanation, counterfactual reasoning, and human-in-the-loop modification---and have strong empirical performance. Experiments over a suite of goal recognition and behaviour classification datasets show our learned automata-based classifiers to have comparable test performance to LSTM-based classifiers, with the added advantage of being interpretable.

Patient representation learning refers to learning a dense mathematical representation of a patient that encodes meaningful information from Electronic Health Records (EHRs). This is generally performed using advanced deep learning methods. This study presents a systematic review of this field and provides both qualitative and quantitative analyses from a methodological perspective. We identified studies developing patient representations from EHRs with deep learning methods from MEDLINE, EMBASE, Scopus, the Association for Computing Machinery (ACM) Digital Library, and Institute of Electrical and Electronics Engineers (IEEE) Xplore Digital Library. After screening 362 articles, 48 papers were included for a comprehensive data collection. We noticed a typical workflow starting with feeding raw data, applying deep learning models, and ending with clinical outcome predictions as evaluations of the learned representations. Specifically, learning representations from structured EHR data was dominant (36 out of 48 studies). Recurrent Neural Networks were widely applied as the deep learning architecture (LSTM: 13 studies, GRU: 11 studies). Disease prediction was the most common application and evaluation (30 studies). Benchmark datasets were mostly unavailable (28 studies) due to privacy concerns of EHR data, and code availability was assured in 20 studies. We show the importance and feasibility of learning comprehensive representations of patient EHR data through a systematic review. Advances in patient representation learning techniques will be essential for powering patient-level EHR analyses. Future work will still be devoted to leveraging the richness and potential of available EHR data. Knowledge distillation and advanced learning techniques will be exploited to assist the capability of learning patient representation further.

Canonical morphological segmentation consists of dividing words into their standardized morphemes. Here, we are interested in approaches for the task when training data is limited. We compare model performance in a simulated low-resource setting for the high-resource languages German, English, and Indonesian to experiments on new datasets for the truly low-resource languages Popoluca and Tepehua. We explore two new models for the task, borrowing from the closely related area of morphological generation: an LSTM pointer-generator and a sequence-to-sequence model with hard monotonic attention trained with imitation learning. We find that, in the low-resource setting, the novel approaches outperform existing ones on all languages by up to 11.4% accuracy. However, while accuracy in emulated low-resource scenarios is over 50% for all languages, for the truly low-resource languages Popoluca and Tepehua, our best model only obtains 37.4% and 28.4% accuracy, respectively. Thus, we conclude that canonical segmentation is still a challenging task for low-resource languages.