Suffix Bidirectional LSTM (Suffix BiLSTM)
Recurrent neural networks have become ubiquitous in computing representations of sequential data, especially textual data in natural language processing. In particular, Bidirectional LSTMs are at the heart of several neural models achieving state-of-the-art performance in a wide variety of tasks in NLP. We propose a general and effective improvement to the BiLSTM model which encodes each suffix and prefix of a sequence of tokens in both forward and reverse directions. We call our model Suffix BiLSTM or SuBiLSTM. Using an extensive set of experiments, we demonstrate that using SuBiLSTM instead of a BiLSTM in existing base models leads to improvements in performance in learning general sentence representations, text classification, textual entailment and named entity recognition. We achieve new state-of-the-art results for fine-grained sentiment classification and question classification using SuBiLSTM. …

Localized System Solver (locSolver)
Column generation is often used to solve multi-commodity flow problems. A program for column generation always includes a module that solves a linear equation. In this paper, we address three major issues in solving linear problem during column generation procedure which are (1) how to employ the sparse property of the coefficient matrix; (2) how to reduce the size of the coefficient matrix; and (3) how to reuse the solution to a similar equation. To this end, we first analyze the sparse property of coefficient matrix of linear equations and find that the matrices occurring in iteration are very sparse. Then, we present an algorithm locSolver (for localized system solver) for linear equations with sparse coefficient matrices and right-hand-sides. This algorithm can reduce the number of variables. After that, we present the algorithm incSolver (for incremental system solver) which utilizes similarity in the iterations of the program for a linear equation system. All three techniques can be used in column generation of multi-commodity problems. Preliminary numerical experiments show that the incSolver is significantly faster than the existing algorithms. For example, random test cases show that incSolver is at least 37 times and up to 341 times faster than popular solver LAPACK. …

DEFRAG
Extreme classification seeks to assign each data point, the most relevant labels from a universe of a million or more labels. This task is faced with the dual challenge of high precision and scalability, with millisecond level prediction times being a benchmark. We propose DEFRAG, an adaptive feature agglomeration technique to accelerate extreme classification algorithms. Despite past works on feature clustering and selection, DEFRAG distinguishes itself in being able to scale to millions of features, and is especially beneficial when feature sets are sparse, which is typical of recommendation and multi-label datasets. The method comes with provable performance guarantees and performs efficient task-driven agglomeration to reduce feature dimensionalities by an order of magnitude or more. Experiments show that DEFRAG can not only reduce training and prediction times of several leading extreme classification algorithms by as much as 40%, but also be used for feature reconstruction to address the problem of missing features, as well as offer superior coverage on rare labels. …

Constrained Optimisation With Latent Distribution (COLD)
Optimising discrete data for a desired characteristic using gradient-based methods involves projecting the data into a continuous latent space and carrying out optimisation in this space. Carrying out global optimisation is difficult as optimisers are likely to follow gradients into regions of the latent space that the model has not been exposed to during training; samples generated from these regions are likely to be too dissimilar to the training data to be useful. We propose Constrained Optimisation with Latent Distributions (COLD), a constrained global optimisation procedure to find samples with high values of a desired property that are similar to yet distinct from the training data. We find that on MNIST, our procedure yields optima for each of three different objectives, and that enforcing tighter constraints improves the quality and increases the diversity of the generated images. On the ChEMBL molecular dataset, our method generates a diverse set of new molecules with drug-likeness scores similar to those of the highest-scoring molecules in the training data. We also demonstrate a computationally efficient way to approximate the constraint when evaluating it exactly is computationally expensive. …