This paper explores a simple and efficient baseline for text classification. Our experiments show that our fast text classifier fastText is often on par with deep learning classifiers in terms of accuracy, and many orders of magnitude faster for training and evaluation. We can train fastText on more than one billion words in less than ten minutes using a standard multicore~CPU, and classify half a million sentences among~312K classes in less than a minute.
When working with three-dimensional data, choice of representation is key. We explore voxel-based models, and present evidence for the viability of voxellated representations in applications including shape modeling and object classification. Our key contributions are methods for training voxel-based variational autoencoders, a user interface for exploring the latent space learned by the autoencoder, and a deep convolutional neural network architecture for object classification. We address challenges unique to voxel-based representations, and empirically evaluate our models on the ModelNet benchmark, where we demonstrate a 51.5% relative improvement in the state of the art for object classification.
We develop an automated variational method for inference in models with Gaussian process (GP) priors and general likelihoods. The method supports multiple outputs and multiple latent functions and does not require detailed knowledge of the conditional likelihood, only needing its evaluation as a black-box function. Using a mixture of Gaussians as the variational distribution, we show that the evidence lower bound and its gradients can be estimated efficiently using samples from univariate Gaussian distributions. Furthermore, the method is scalable to large datasets which is achieved by using an augmented prior via the inducing-variable approach underpinning most sparse GP approximations, along with parallel computation and stochastic optimization. We evaluate our approach quantitatively and qualitatively with experiments on small datasets, medium-scale datasets and large datasets, showing its competitiveness under different likelihood models and sparsity levels. On the large-scale experiments involving prediction of airline delays and classification of handwritten digits, we show that our method is on par with the state-of-the-art hard-coded approaches for scalable GP regression and classification.
Image classification is one of the main research problems in computer vision and machine learning. Since in most real-world image classification applications there is no control over how the images are captured, it is necessary to consider the possibility that these images might be affected by noise (e.g. sensor noise in a low-quality surveillance camera). In this paper we analyse the impact of three different types of noise on descriptors extracted by two widely used feature extraction methods (LBP and HOG) and how denoising the images can help to mitigate this problem. We carry out experiments on two different datasets and consider several types of noise, noise levels, and denoising methods. Our results show that noise can hinder classification performance considerably and make classes harder to separate. Although denoising methods were not able to reach the same performance of the noise-free scenario, they improved classification results for noisy data.
Several recent deep neural networks experiments leverage the generalist-specialist paradigm for classification. However, no formal study compared the performance of different clustering algorithms for class assignment. In this paper we perform such a study, suggest slight modifications to the clustering procedures, and propose a novel algorithm designed to optimize the performance of of the specialist-generalist classification system. Our experiments on the CIFAR-10 and CIFAR-100 datasets allow us to investigate situations for varying number of classes on similar data. We find that our \emph{greedy pairs} clustering algorithm consistently outperforms other alternatives, while the choice of the confusion matrix has little impact on the final performance.
Stochastic gradient-boosted decision trees are widely employed for multivariate classification and regression tasks. This paper presents a speed-optimized and cache-friendly implementation for multivariate classification called FastBDT. FastBDT is one order of magnitude faster during the fitting-phase and application-phase, in comparison with popular implementations in software frameworks like TMVA, scikit-learn and XGBoost. The concepts used to optimize the execution time and performance studies are discussed in detail in this paper. The key ideas include: An equal-frequency binning on the input data, which allows replacing expensive floating-point with integer operations, while at the same time increasing the quality of the classification; a cache-friendly linear access pattern to the input data, in contrast to usual implementations, which exhibit a random access pattern. FastBDT provides interfaces to C/C++, Python and TMVA. It is extensively used in the field of high energy physics by the Belle II experiment.
The staggering amount of streaming time series coming from the real world calls for more efficient and effective online modeling solution. For time series modeling, most existing works make some unrealistic assumptions such as the input data is of fixed length or well aligned, which requires extra effort on segmentation or normalization of the raw streaming data. Although some literature claim their approaches to be invariant to data length and misalignment, they are too time-consuming to model a streaming time series in an online manner. We propose a novel and more practical online modeling and classification scheme, DDE-MGM, which does not make any assumptions on the time series while maintaining high efficiency and state-of-the-art performance. The derivative delay embedding (DDE) is developed to incrementally transform time series to the embedding space, where the intrinsic characteristics of data is preserved as recursive patterns regardless of the stream length and misalignment. Then, a non-parametric Markov geographic model (MGM) is proposed to both model and classify the pattern in an online manner. Experimental results demonstrate the effectiveness and superior classification accuracy of the proposed DDE-MGM in an online setting as compared to the state-of-the-art.
We present a new Convolutional Neural Network (CNN) model for text classification that jointly exploits labels on documents and their component sentences. Specifically, we consider scenarios in which annotators explicitly mark sentences (or snippets) that support their overall document categorization, i.e., they provide rationales. Our model exploits such supervision via a hierarchical approach in which each document is represented by a linear combination of the vector representations of its component sentences. We propose a sentence-level convolutional model that estimates the probability that a given sentence is a rationale, and we then scale the contribution of each sentence to the aggregate document representation in proportion to these estimates. Experiments on five classification datasets that have document labels and associated rationales demonstrate that our approach consistently outperforms strong baselines. Moreover, our model naturally provides explanations for its predictions.
Deep learning has established many new state of the art solutions in the last decade in areas such as object, scene and speech recognition. In particular Convolutional Neural Network (CNN) is a category of deep learning which obtains excellent results in object detection and recognition tasks. Its architecture is indeed well suited to object analysis by learning and classifying complex (deep) features that represent parts of an object or the object itself. However, some of its features are very similar to texture analysis methods. CNN layers can be thought of as filter banks of complexity increasing with the depth. Filter banks are powerful tools to extract texture features and have been widely used in texture analysis. In this paper we develop a simple network architecture named Texture CNN (T-CNN) which explores this observation. It is built on the idea that the overall shape information extracted by the fully connected layers of a classic CNN is of minor importance in texture analysis. Therefore, we pool an energy measure from the last convolution layer which we connect to a fully connected layer. We show that our approach can improve the performance of a network while greatly reducing the memory usage and computation.
The main objective of this project is to segment different breast ultrasound images to find out lesion area by discarding the low contrast regions as well as the inherent speckle noise. The proposed method consists of three stages (removing noise, segmentation, classification) in order to extract the correct lesion. We used normalized cuts approach to segment ultrasound images into regions of interest where we can possibly finds the lesion, and then K-means classifier is applied to decide finally the location of the lesion. For every original image, an annotated ground-truth image is given to perform comparison with the obtained experimental results, providing accurate evaluation measures.
Systems based on artificial neural networks (ANNs) have achieved state-of-the-art results in many natural language processing tasks. Although ANNs do not require manually engineered features, ANNs have many hyperparameters to be optimized. The choice of hyperparameters significantly impacts models' performances. However, the ANN hyperparameters are typically chosen by manual, grid, or random search, which either requires expert experiences or is computationally expensive. Recent approaches based on Bayesian optimization using Gaussian processes (GPs) is a more systematic way to automatically pinpoint optimal or near-optimal machine learning hyperparameters. Using a previously published ANN model yielding state-of-the-art results for dialog act classification, we demonstrate that optimizing hyperparameters using GP further improves the results, and reduces the computational time by a factor of 4 compared to a random search. Therefore it is a useful technique for tuning ANN models to yield the best performances for natural language processing tasks.
Images are formed by counting how many photons traveling from a given set of directions hit an image sensor during a given time interval. When photons are few and far in between, the concept of `image' breaks down and it is best to consider directly the flow of photons. Computer vision in this regime, which we call `scotopic', is radically different from the classical image-based paradigm in that visual computations (classification, control, search) have to take place while the stream of photons is captured and decisions may be taken as soon as enough information is available. The scotopic regime is important for biomedical imaging, security, astronomy and many other fields. Here we develop a framework that allows a machine to classify objects with as few photons as possible, while maintaining the error rate below an acceptable threshold. A dynamic and asymptotically optimal speed-accuracy tradeoff is a key feature of this framework. We propose and study an algorithm to optimize the tradeoff of a convolutional network directly from lowlight images and evaluate on simulated images from standard datasets. Surprisingly, scotopic systems can achieve comparable classification performance as traditional vision systems while using less than 0.1% of the photons in a conventional image. In addition, we demonstrate that our algorithms work even when the illuminance of the environment is unknown and varying. Last, we outline a spiking neural network coupled with photon-counting sensors as a power-efficient hardware realization of scotopic algorithms.
Convolutional Neural Networks (CNNs) have gained significant traction in the field of machine learning, particularly due to their high accuracy in visual recognition. Recent works have pushed the performance of GPU implementations of CNNs to significantly improve their classification and training times. With these improvements, many frameworks have become available for implementing CNNs on both CPUs and GPUs, with no support for FPGA implementations. In this work we present a modified version of the popular CNN framework Caffe, with FPGA support. This allows for classification using CNN models and specialized FPGA implementations with the flexibility of reprogramming the device when necessary, seamless memory transactions between host and device, simple-to-use test benches, and the ability to create pipelined layer implementations. To validate the framework, we use the Xilinx SDAccel environment to implement an FPGA-based Winograd convolution engine and show that the FPGA layer can be used alongside other layers running on a host processor to run several popular CNNs (AlexNet, GoogleNet, VGG A, Overfeat). The results show that our framework achieves 50 GFLOPS across 3x3 convolutions in the benchmarks. This is achieved within a practical framework, which will aid in future development of FPGA-based CNNs.
This paper presents how we can achieve the state-of-the-art accuracy in multi-category object detection task while minimizing the computational cost by adapting and combining recent technical innovations. Following the common pipeline of "CNN feature extraction + region proposal + RoI classification", we mainly redesign the feature extraction part, since region proposal part is not computationally expensive and classification part can be efficiently compressed with common techniques like truncated SVD. Our design principle is "less channels with more layers" and adoption of some building blocks including concatenated ReLU, Inception, and HyperNet. The designed network is deep and thin and trained with the help of batch normalization, residual connections, and learning rate scheduling based on plateau detection. We obtained solid results on well-known object detection benchmarks: 83.8% mAP (mean average precision) on VOC2007 and 82.5% mAP on VOC2012 (2nd place), while taking only 750ms/image on Intel i7-6700K CPU with a single core and 46ms/image on NVIDIA Titan X GPU. Theoretically, our network requires only 12.3% of the computational cost compared to ResNet-101, the winner on VOC2012.
Neural networks are known to be effective function approximators. Recently, deep neural networks have proven to be very effective in pattern recognition, classification tasks and human-level control to model highly nonlinear realworld systems. This paper investigates the effectiveness of deep neural networks in the modeling of dynamical systems with complex behavior. Three deep neural network structures are trained on sequential data, and we investigate the effectiveness of these networks in modeling associated characteristics of the underlying dynamical systems. We carry out similar evaluations on select publicly available system identification datasets. We demonstrate that deep neural networks are effective model estimators from input-output data
Most existing star-galaxy classifiers use the reduced summary information from catalogs, requiring careful feature extraction and selection. The latest advances in machine learning that use deep convolutional neural networks allow a machine to automatically learn the features directly from data, minimizing the need for input from human experts. We present a star-galaxy classification framework that uses deep convolutional neural networks (ConvNets) directly on the reduced, calibrated pixel values. Using data from the Sloan Digital Sky Survey (SDSS) and the Canada-France-Hawaii Telescope Lensing Survey (CFHTLenS), we demonstrate that ConvNets are able to produce accurate and well-calibrated probabilistic classifications that are competitive with conventional machine learning techniques. Future advances in deep learning may bring more success with current and forthcoming photometric surveys, such as the Dark Energy Survey (DES) and the Large Synoptic Survey Telescope (LSST), because deep neural networks require very little, manual feature engineering.
We investigate metric learning in the context of dynamic time warping (DTW), the by far most popular dissimilarity measure used for the comparison and analysis of motion capture data. While metric learning enables a problem-adapted representation of data, the majority of methods has been proposed for vectorial data only. In this contribution, we extend the popular principle offered by the large margin nearest neighbors learner (LMNN) to DTW by treating the resulting component-wise dissimilarity values as features. We demonstrate that this principle greatly enhances the classification accuracy in several benchmarks. Further, we show that recent auxiliary concepts such as metric regularization can be transferred from the vectorial case to component-wise DTW in a similar way. We illustrate that metric regularization constitutes a crucial prerequisite for the interpretation of the resulting relevance profiles.
Generating user interpretable multi-class predictions in data rich environments with many classes and explanatory covariates is a daunting task. We introduce Diagonal Orthant Latent Dirichlet Allocation (DOLDA), a supervised topic model for multi-class classification that can handle both many classes as well as many covariates. To handle many classes we use the recently proposed Diagonal Orthant (DO) probit model (Johndrow et al., 2013) together with an efficient Horseshoe prior for variable selection/shrinkage (Carvalho et al., 2010). We propose a computationally efficient parallel Gibbs sampler for the new model. An important advantage of DOLDA is that learned topics are directly connected to individual classes without the need for a reference class. We evaluate the model's predictive accuracy on two datasets and demonstrate DOLDA's advantage in interpreting the generated predictions.
We introduce word vectors for the construction domain. Our vectors were obtained by running word2vec on an 11M-word corpus that we created from scratch by leveraging freely-accessible online sources of construction-related text. We first explore the embedding space and show that our vectors capture meaningful construction-specific concepts. We then evaluate the performance of our vectors against that of ones trained on a 100B-word corpus (Google News) within the framework of an injury report classification task. Without any parameter tuning, our embeddings give competitive results, and outperform the Google News vectors in many cases. Using a keyword-based compression of the reports also leads to a significant speed-up with only a limited loss in performance. We release our corpus and the data set we created for the classification task as publicly available, in the hope that they will be used by future studies for benchmarking and building on our work.
In this paper we propose a novel approach for learning from data using rule based fuzzy inference systems where the model parameters are estimated using Bayesian inference and Markov Chain Monte Carlo (MCMC) techniques. We show the applicability of the method for regression and classification tasks using synthetic data-sets and also a real world example in the financial services industry. Then we demonstrate how the method can be extended for knowledge extraction to select the individual rules in a Bayesian way which best explains the given data. Finally we discuss the advantages and pitfalls of using this method over state-of-the-art techniques and highlight the specific class of problems where this would be useful.