We propose an approach to decomposing a thematic information stream into principal components. Each principal component is related to a narrow topic extracted from the information stream. The essence of the approach arises from analogy with the Fourier transform. We examine methods for analyzing the principal components and propose using multifractal analysis for identifying similar topics. The decomposition technique is applied to the information stream dedicated to Brexit. We provide a comparison between the principal components obtained by applying the decomposition to Brexit stream and the related topics extracted by Google Trends.
Autoencoders learn data representations (codes) in such a way that the input is reproduced at the output of the network. However, it is not always clear what kind of properties of the input data need to be captured by the codes. Kernel machines have experienced great success by operating via inner-products in a theoretically well-defined reproducing kernel Hilbert space, hence capturing topological properties of input data. In this paper, we enhance the autoencoder’s ability to learn effective data representations by aligning inner products between codes with respect to a kernel matrix. By doing so, the proposed kernelized autoencoder allows learning similarity-preserving embeddings of input data, where the notion of similarity is explicitly controlled by the user and encoded in a positive semi-definite kernel matrix. Experiments are performed for evaluating both reconstruction and kernel alignment performance in classification tasks and visualization of high-dimensional data. Additionally, we show that our method is capable to emulate kernel principal component analysis on a denoising task, obtaining competitive results at a much lower computational cost.
Any autoencoder network can be turned into a generative model by imposing an arbitrary prior distribution on its hidden code vector. Variational Autoencoder (VAE) [2] uses a KL divergence penalty to impose the prior, whereas Adversarial Autoencoder (AAE) [1] uses {\it generative adversarial networks} GAN [3]. GAN trades the complexities of {\it sampling} algorithms with the complexities of {\it searching} Nash equilibrium in minimax games. Such minimax architectures get trained with the help of data examples and gradients flowing through a generator and an adversary. A straightforward modification of AAE is to replace the adversary with the maximum mean discrepancy (MMD) test [4-5]. This replacement leads to a new type of probabilistic autoencoder, which is also discussed in our paper. We propose a novel probabilistic autoencoder in which the adversary of AAE is replaced with a space of {\it stochastic} functions. This replacement introduces a new source of randomness, which can be considered as a continuous control for encouraging {\it explorations}. This prevents the adversary from fitting too closely to the generator and therefore leads to a more diverse set of generated samples.
Inspired by Kalikow-type decompositions, we introduce a new stochastic model of infinite neuronal networks, for which we establish oracle inequalities for Lasso methods and restricted eigenvalue properties for the associated Gram matrix with high probability. These results hold even if the network is only partially observed. The main argument rely on the fact that concentration inequalities can easily be derived whenever the transition probabilities of the underlying process admit a sparse space-time representation.
Reservoir computing is a neural network approach for processing time-dependent signals that has seen rapid development in recent years. Physical implementations of the technique using optical reservoirs have demonstrated remarkable accuracy and processing speed at benchmark tasks. However, these approaches require an electronic output layer to maintain high performance, which limits their use in tasks such as time-series prediction, where the output is fed back into the reservoir. We present here a reservoir computing scheme that has rapid processing speed both by the reservoir and the output layer. The reservoir is realized by an autonomous, time-delay, Boolean network configured on a field-programmable gate array. We investigate the dynamical properties of the network and observe the fading memory property that is critical for successful reservoir computing. We demonstrate the utility of the technique by training a reservoir to learn the short- and long-term behavior of a chaotic system. We find accuracy comparable to state-of-the-art software approaches of similar network size, but with a superior real-time prediction rate up to 160 MHz.
Resource Description Framework (RDF) has been widely used to represent information on the web, while SPARQL is a standard query language to manipulate RDF data. Given a SPARQL query, there often exist many joins which are the bottlenecks of efficiency of query processing. Besides, the real RDF datasets often reveal strong data sparsity, which indicates that a resource often only relates to a few resources even the number of total resources is large. In this paper, we propose a sparse matrix-based (SM-based) SPARQL query processing approach over RDF datasets which con- siders both join optimization and data sparsity. Firstly, we present a SM-based storage for RDF datasets to lift the storage efficiency, where valid edges are stored only, and then introduce a predicate- based hash index on the storage. Secondly, we develop a scalable SM-based join algorithm for SPARQL query processing. Finally, we analyze the overall cost by accumulating all intermediate results and design a query plan generated algorithm. Besides, we extend our SM-based join algorithm on GPU for parallelizing SPARQL query processing. We have evaluated our approach compared with the state-of-the-art RDF engines over benchmark RDF datasets and the experimental results show that our proposal can significantly improve SPARQL query processing with high scalability.
Studies have demonstrated that Apache Spark, Flink and related frameworks can perform stream processing at very high frequencies, whilst tending to focus on small messages with a computationally light `map’ stage for each message; a common enterprise use case. We add to these benchmarks by broadening the domain to include loads with larger messages (leading to network-bound throughput), and that are computationally intensive (leading to CPU-bound throughput) in the map phase; in order to evaluate applicability of these frameworks to scientific computing applications. We present a performance benchmark comparison between Apache Spark Streaming (ASS) under both file and TCP streaming modes; and HarmonicIO, comparing maximum throughput over a broad domain of message sizes and CPU loads. We find that relative performance varies considerably across this domain, with the chosen means of stream source integration having a big impact. We offer recommendations for choosing and configuring the frameworks, and present a benchmarking toolset developed for this study.
Employers actively look for talents having not only specific hard skills but also various soft skills. To analyze the soft skill demands on the job market, it is important to be able to detect soft skill phrases from job advertisements automatically. However, a naive matching of soft skill phrases can lead to false positive matches when a soft skill phrase, such as friendly, is used to describe a company, a team, or another entity, rather than a desired candidate. In this paper, we propose a phrase-matching-based approach which differentiates between soft skill phrases referring to a candidate vs. something else. The disambiguation is formulated as a binary text classification problem where the prediction is made for the potential soft skill based on the context where it occurs. To inform the model about the soft skill for which the prediction is made, we develop several approaches, including soft skill masking and soft skill tagging. We compare several neural network based approaches, including CNN, LSTM and Hierarchical Attention Model. The proposed tagging-based input representation using LSTM achieved the highest recall of 83.92% on the job dataset when fixing a precision to 95%.
Social media is increasingly used by humans to express their feelings and opinions in the form of short text messages. Detecting sentiments in the text has a wide range of applications including identifying anxiety or depression of individuals and measuring well-being or mood of a community. Sentiments can be expressed in many ways that can be seen such as facial expression and gestures, speech and by written text. Sentiment Analysis in text documents is essentially a content-based classification problem involving concepts from the domains of Natural Language Processing as well as Machine Learning. In this paper, sentiment recognition based on textual data and the techniques used in sentiment analysis are discussed.
Named entities (NE) are objects that are referred to by names such as people, organizations and locations. Named entities and keywords are important to the meaning of a document. We propose a generalized vector space model that combines named entities and keywords. In the model, we take into account different ontological features of named entities, namely, aliases, classes and identifiers. Moreover, we use entity classes to represent the latent information of interrogative words in Wh-queries, which are ignored in traditional keyword-based searching. We have implemented and tested the proposed model on a TREC dataset, as presented and discussed in the paper.
Similarity and metric learning provides a principled approach to construct a task-specific similarity from weakly supervised data. However, these methods are subject to the curse of dimensionality: as the number of features grows large, poor generalization is to be expected and training becomes intractable due to high computational and memory costs. In this paper, we propose a similarity learning method that can efficiently deal with high-dimensional sparse data. This is achieved through a parameterization of similarity functions by convex combinations of sparse rank-one matrices, together with the use of a greedy approximate Frank-Wolfe algorithm which provides an efficient way to control the number of active features. We show that the convergence rate of the algorithm, as well as its time and memory complexity, are independent of the data dimension. We further provide a theoretical justification of our modeling choices through an analysis of the generalization error, which depends logarithmically on the sparsity of the solution rather than on the number of features. Our experiments on datasets with up to one million features demonstrate the ability of our approach to generalize well despite the high dimensionality as well as its superiority compared to several competing methods.
This paper introduces a new tool for time-series analysis: the Sliding Window Discrete Fourier Transform (SWDFT). The SWDFT is especially useful for time-series with local- in-time periodic components. We define a 5-parameter model for noiseless local periodic signals, then study the SWDFT of this model. Our study illustrates several key concepts crucial to analyzing time-series with the SWDFT, in particular Aliasing, Leakage, and Ringing. We also show how these ideas extend to R > 1 local periodic components, using the linearity property of the Fourier transform. Next, we propose a simple procedure for estimating the 5 parameters of our local periodic signal model using the SWDFT. Our estimation procedure speeds up computation by using a trigonometric identity that linearizes estimation of 2 of the 5 parameters. We conclude with a very small Monte Carlo simulation study of our estimation procedure under different levels of noise.
Increased information sharing through short and long-range skip connections between layers in fully convolutional networks have demonstrated significant improvement in performance for semantic segmentation. In this paper, we propose Competitive Dense Fully Convolutional Networks (CDFNet) by introducing competitive maxout activations in place of naive feature concatenation for inducing competition amongst layers. Within CDFNet, we propose two architectural contributions, namely competitive dense block (CDB) and competitive unpooling block (CUB) to induce competition at local and global scales for short and long-range skip connections respectively. This extension is demonstrated to boost learning of specialized sub-networks targeted at segmenting specific anatomies, which in turn eases the training of complex tasks. We present the proof-of-concept on the challenging task of whole body segmentation in the publicly available VISCERAL benchmark and demonstrate improved performance over multiple learning and registration based state-of-the-art methods.
Interactive massively parallel computations are critical for machine learning and data analysis. These computations are a staple of the MIT Lincoln Laboratory Supercomputing Center (LLSC) and has required the LLSC to develop unique interactive supercomputing capabilities. Scaling interactive machine learning frameworks, such as TensorFlow, and data analysis environments, such as MATLAB/Octave, to tens of thousands of cores presents many technical challenges – in particular, rapidly dispatching many tasks through a scheduler, such as Slurm, and starting many instances of applications with thousands of dependencies. Careful tuning of launches and prepositioning of applications overcome these challenges and allow the launching of thousands of tasks in seconds on a 40,000-core supercomputer. Specifically, this work demonstrates launching 32,000 TensorFlow processes in 4 seconds and launching 262,000 Octave processes in 40 seconds. These capabilities allow researchers to rapidly explore novel machine learning architecture and data analysis algorithms.
Given two random variables $X$ and $Y$, an operational approach is undertaken to quantify the “leakage” of information from $X$ to $Y$. The resulting measure $\mathcal{L}(X \!\! \to \!\! Y)$ is called \emph{maximal leakage}, and is defined as the multiplicative increase, upon observing $Y$, of the probability of correctly guessing a randomized function of $X$, maximized over all such randomized functions. A closed-form expression for $\mathcal{L}(X \!\! \to \!\! Y)$ is given for discrete $X$ and $Y$, and it is subsequently generalized to handle a large class of random variables. The resulting properties are shown to be consistent with an axiomatic view of a leakage measure, and the definition is shown to be robust to variations in the setup. Moreover, a variant of the Shannon cipher system is studied, in which performance of an encryption scheme is measured using maximal leakage. A single-letter characterization of the optimal limit of (normalized) maximal leakage is derived and asymptotically-optimal encryption schemes are demonstrated. Furthermore, the sample complexity of estimating maximal leakage from data is characterized up to subpolynomial factors. Finally, the \emph{guessing} framework used to define maximal leakage is used to give operational interpretations of commonly used leakage measures, such as Shannon capacity, maximal correlation, and local differential privacy.
This paper presents a novel, causally-inspired approach to domain adaptation which aims to also include unlabelled data in the model fitting when labelled data is scarce. We consider a case of covariate-shift adaptation with cause and effect features, and–drawing from recent ideas in causal modelling and invariant prediction–show how this setting leads to, what we will refer to as, a semi-generative model: P(Y, X_eff; X_cau,\theta). Our proposed approach is robust to changes in the distribution over causal features, and naturally allows to impose model constraints by unsupervised learning of a map from causes to effects. In experiments on synthetic datasets we demonstrate a significant improvement in classification performance of our semi-generative model over purely-supervised and importance-weighting baselines when the amount of labelled data is small. Moreover, we apply our approach for regression on real-world protein-count data and compare it to feature transformation methods.
Uplift modeling is an area of machine learning which aims at predicting the causal effect of some action on a given individual. The action may be a medical procedure, marketing campaign, or any other circumstance controlled by the experimenter. Building an uplift model requires two training sets: the treatment group, where individuals have been subject to the action, and the control group, where no action has been performed. An uplift model allows then to assess the gain resulting from taking the action on a given individual, such as the increase in probability of patient recovery or of a product being purchased. This paper describes an adaptation of the well-known boosting techniques to the uplift modeling case. We formulate three desirable properties which an uplift boosting algorithm should have. Since all three properties cannot be satisfied simultaneously, we propose three uplift boosting algorithms, each satisfying two of them. Experiments demonstrate the usefulness of the proposed methods, which often dramatically improve performance of the base models and are thus new and powerful tools for uplift modeling.
Deep convolution neural network has achieved great success in many artificial intelligence applications. However, its enormous model size and massive computation cost have become the main obstacle for deployment of such powerful algorithm in the low power and resource-limited embedded systems. As the countermeasure to this problem, in this work, we propose statistical weight scaling and residual expansion methods to reduce the bit-width of the whole network weight parameters to ternary values (i.e. -1, 0, +1), with the objectives to greatly reduce model size, computation cost and accuracy degradation caused by the model compression. With about 16x model compression rate, our ternarized ResNet-32/44/56 could outperform full-precision counterparts by 0.12%, 0.24% and 0.18% on CIFAR- 10 dataset. We also test our ternarization method with AlexNet and ResNet-18 on ImageNet dataset, which both achieve the best top-1 accuracy compared to recent similar works, with the same 16x compression rate. If further incorporating our residual expansion method, compared to the full-precision counterpart, our ternarized ResNet-18 even improves the top-5 accuracy by 0.61% and merely degrades the top-1 accuracy only by 0.42% for the ImageNet dataset, with 8x model compression rate. It outperforms the recent ABC-Net by 1.03% in top-1 accuracy and 1.78% in top-5 accuracy, with around 1.25x higher compression rate and more than 6x computation reduction due to the weight sparsity.