Modern compression algorithms are often the result of laborious domain-specific research; industry standards such as MP3, JPEG, and AMR-WB took years to develop and were largely hand-designed. We present a deep neural network model which optimizes all the steps of a wideband speech coding pipeline (compression, quantization, entropy coding, and decompression) end-to-end directly from raw speech data -- no manual feature engineering necessary, and it trains in hours. In testing, our DNN-based coder performs on par with the AMR-WB standard at a variety of bitrates (~9kbps up to ~24kbps). It also runs in realtime on a 3.8GhZ Intel CPU.
Deep reinforcement learning (DRL) methods such as the Deep Q-Network (DQN) have achieved state-of-the-art results in a variety of challenging, high-dimensional domains. This success is mainly attributed to the power of deep neural networks to learn rich domain representations for approximating the value function or policy. Batch reinforcement learning methods with linear representations, on the other hand, are more stable and require less hyper parameter tuning. Yet, substantial feature engineering is necessary to achieve good results. In this work we propose a hybrid approach -- the Least Squares Deep Q-Network (LS-DQN), which combines rich feature representations learned by a DRL algorithm with the stability of a linear least squares method. We do this by periodically re-training the last hidden layer of a DRL network with a batch least squares update. Key to our approach is a Bayesian regularization term for the least squares update, which prevents over-fitting to the more recent data. We tested LS-DQN on five Atari games and demonstrate significant improvement over vanilla DQN and Double-DQN. We also investigated the reasons for the superior performance of our method. Interestingly, we found that the performance improvement can be attributed to the large batch size used by the LS method when optimizing the last layer.
Character-based sequence labeling framework is flexible and efficient for Chinese word segmentation (CWS). Recently, many character-based neural models have been applied to CWS. While they obtain good performance, they have two obvious weaknesses. The first is that they heavily rely on manually designed bigram feature, i.e. they are not good at capturing n-gram features automatically. The second is that they make no use of full word information. For the first weakness, we propose a convolutional neural model, which is able to capture rich n-gram features without any feature engineering. For the second one, we propose an effective approach to integrate the proposed model with word embeddings. We evaluate the model on two benchmark datasets: PKU and MSR. Without any feature engineering, the model obtains competitive performance -- 95.7% on PKU and 97.3% on MSR. Armed with word embeddings, the model achieves state-of-the-art performance on both datasets -- 96.5% on PKU and 98.0% on MSR, without using any external labeled resource.
Deep learning has demonstrated tremendous potential for Automatic Text Scoring (ATS) tasks. In this paper, we describe a new neural architecture that enhances vanilla neural network models with auxiliary neural coherence features. Our new method proposes a new \textsc{SkipFlow} mechanism that models relationships between snapshots of the hidden representations of a long short-term memory (LSTM) network as it reads. Subsequently, the semantic relationships between multiple snapshots are used as auxiliary features for prediction. This has two main benefits. Firstly, essays are typically long sequences and therefore the memorization capability of the LSTM network may be insufficient. Implicit access to multiple snapshots can alleviate this problem by acting as a protection against vanishing gradients. The parameters of the \textsc{SkipFlow} mechanism also acts as an auxiliary memory. Secondly, modeling relationships between multiple positions allows our model to learn features that represent and approximate textual coherence. In our model, we call this \textit{neural coherence} features. Overall, we present a unified deep learning architecture that generates neural coherence features as it reads in an end-to-end fashion. Our approach demonstrates state-of-the-art performance on the benchmark ASAP dataset, outperforming not only feature engineering baselines but also other deep learning models.
Accurate detection of objects in 3D point clouds is a central problem in many applications, such as autonomous navigation, housekeeping robots, and augmented/virtual reality. To interface a highly sparse LiDAR point cloud with a region proposal network (RPN), most existing efforts have focused on hand-crafted feature representations, for example, a bird's eye view projection. In this work, we remove the need of manual feature engineering for 3D point clouds and propose VoxelNet, a generic 3D detection network that unifies feature extraction and bounding box prediction into a single stage, end-to-end trainable deep network. Specifically, VoxelNet divides a point cloud into equally spaced 3D voxels and transforms a group of points within each voxel into a unified feature representation through the newly introduced voxel feature encoding (VFE) layer. In this way, the point cloud is encoded as a descriptive volumetric representation, which is then connected to a RPN to generate detections. Experiments on the KITTI car detection benchmark show that VoxelNet outperforms the state-of-the-art LiDAR based 3D detection methods by a large margin. Furthermore, our network learns an effective discriminative representation of objects with various geometries, leading to encouraging results in 3D detection of pedestrians and cyclists, based on only LiDAR.
Electroencephalography (EEG) is an extensively-used and well-studied technique in the field of medical diagnostics and treatment for brain disorders, including epilepsy, migraines, and tumors. The analysis and interpretation of EEGs require physicians to have specialized training, which is not common even among most doctors in the developed world, let alone the developing world where physician shortages plague society. This problem can be addressed by teleEEG that uses remote EEG analysis by experts or by local computer processing of EEGs. However, both of these options are prohibitively expensive and the second option requires abundant computing resources and infrastructure, which is another concern in developing countries where there are resource constraints on capital and computing infrastructure. In this work, we present a cloud-based deep neural network approach to provide decision support for non-specialist physicians in EEG analysis and interpretation. Named `neurology-as-a-service,' the approach requires almost no manual intervention in feature engineering and in the selection of an optimal architecture and hyperparameters of the neural network. In this study, we deploy a pipeline that includes moving EEG data to the cloud and getting optimal models for various classification tasks. Our initial prototype has been tested only in developed world environments to-date, but our intention is to test it in developing world environments in future work. We demonstrate the performance of our proposed approach using the BCI2000 EEG MMI dataset, on which our service attains 63.4% accuracy for the task of classifying real vs. imaginary activity performed by the subject, which is significantly higher than what is obtained with a shallow approach such as support vector machines.
The dominant neural architectures in question answer retrieval are based on recurrent or convolutional encoders configured with complex word matching layers. Given that recent architectural innovations are mostly new word interaction layers or attention-based matching mechanisms, it seems to be a well-established fact that these components are mandatory for good performance. Unfortunately, the memory and computation cost incurred by these complex mechanisms are undesirable for practical applications. As such, this paper tackles the question of whether it is possible to achieve competitive performance with simple neural architectures. We propose a simple but novel deep learning architecture for fast and efficient question-answer ranking and retrieval. More specifically, our proposed model, \textsc{HyperQA}, is a parameter efficient neural network that outperforms other parameter intensive models such as Attentive Pooling BiLSTMs and Multi-Perspective CNNs on multiple QA benchmarks. The novelty behind \textsc{HyperQA} is a pairwise ranking objective that models the relationship between question and answer embeddings in Hyperbolic space instead of Euclidean space. This empowers our model with a self-organizing ability and enables automatic discovery of latent hierarchies while learning embeddings of questions and answers. Our model requires no feature engineering, no similarity matrix matching, no complicated attention mechanisms nor over-parameterized layers and yet outperforms and remains competitive to many models that have these functionalities on multiple benchmarks.
We present our preliminary work to determine if patient's vocal acoustic, linguistic, and facial patterns could predict clinical ratings of depression severity, namely Patient Health Questionnaire depression scale (PHQ-8). We proposed a multi modal fusion model that combines three different modalities: audio, video , and text features. By training over AVEC 2017 data set, our proposed model outperforms each single modality prediction model, and surpasses the data set baseline with ice margin.
Learning algorithms for natural language processing (NLP) tasks traditionally rely on manually defined relevant contextual features. On the other hand, neural network models using an only distributional representation of words have been successfully applied for several NLP tasks. Such models learn features automatically and avoid explicit feature engineering. Across several domains, neural models become a natural choice specifically when limited characteristics of data are known. However, this flexibility comes at the cost of interpretability. In this paper, we define three different methods to investigate ability of bi-directional recurrent neural networks (RNNs) in capturing contextual features. In particular, we analyze RNNs for sequence tagging tasks. We perform a comprehensive analysis on general as well as biomedical domain datasets. Our experiments focus on important contextual words as features, which can easily be extended to analyze various other feature types. We also investigate positional effects of context words and show how the developed methods can be used for error analysis.
Automated analysis methods are crucial aids for monitoring and defending a network to protect the sensitive or confidential data it hosts. This work introduces a flexible, powerful, and unsupervised approach to detecting anomalous behavior in computer and network logs, one that largely eliminates domain-dependent feature engineering employed by existing methods. By treating system logs as threads of interleaved "sentences" (event log lines) to train online unsupervised neural network language models, our approach provides an adaptive model of normal network behavior. We compare the effectiveness of both standard and bidirectional recurrent neural network language models at detecting malicious activity within network log data. Extending these models, we introduce a tiered recurrent architecture, which provides context by modeling sequences of users' actions over time. Compared to Isolation Forest and Principal Components Analysis, two popular anomaly detection algorithms, we observe superior performance on the Los Alamos National Laboratory Cyber Security dataset. For log-line-level red team detection, our best performing character-based model provides test set area under the receiver operator characteristic curve of 0.98, demonstrating the strong fine-grained anomaly detection performance of this approach on open vocabulary logging sources.
Imitation learning has been commonly applied to solve different tasks in isolation. This usually requires either careful feature engineering, or a significant number of samples. This is far from what we desire: ideally, robots should be able to learn from very few demonstrations of any given task, and instantly generalize to new situations of the same task, without requiring task-specific engineering. In this paper, we propose a meta-learning framework for achieving such capability, which we call one-shot imitation learning. Specifically, we consider the setting where there is a very large set of tasks, and each task has many instantiations. For example, a task could be to stack all blocks on a table into a single tower, another task could be to place all blocks on a table into two-block towers, etc. In each case, different instances of the task would consist of different sets of blocks with different initial states. At training time, our algorithm is presented with pairs of demonstrations for a subset of all tasks. A neural net is trained that takes as input one demonstration and the current state (which initially is the initial state of the other demonstration of the pair), and outputs an action with the goal that the resulting sequence of states and actions matches as closely as possible with the second demonstration. At test time, a demonstration of a single instance of a new task is presented, and the neural net is expected to perform well on new instances of this new task. The use of soft attention allows the model to generalize to conditions and tasks unseen in the training data. We anticipate that by training this model on a much greater variety of tasks and settings, we will obtain a general system that can turn any demonstrations into robust policies that can accomplish an overwhelming variety of tasks. Videos available at https://bit.ly/nips2017-oneshot .
Several studies have shown that the network traffic that is generated by a visit to a website over Tor reveals information specific to the website through the timing and sizes of network packets. By capturing traffic traces between users and their Tor entry guard, a network eavesdropper can leverage this meta-data to reveal which website Tor users are visiting. The success of such attacks heavily depends on the particular set of traffic features that are used to construct the fingerprint. Typically, these features are manually engineered and, as such, any change introduced to the Tor network can render these carefully constructed features ineffective. In this paper, we show that an adversary can automate the feature engineering process, and thus automatically deanonymize Tor traffic by applying our novel method based on deep learning. We collect a dataset comprised of more than three million network traces, which is the largest dataset of web traffic ever used for website fingerprinting, and find that the performance achieved by our deep learning approaches is comparable to known methods which include various research efforts spanning over multiple years. The obtained success rate exceeds 96% for a closed world of 100 websites and 94% for our biggest closed world of 900 classes. In our open world evaluation, the most performant deep learning model is 2% more accurate than the state-of-the-art attack. Furthermore, we show that the implicit features automatically learned by our approach are far more resilient to dynamic changes of web content over time. We conclude that the ability to automatically construct the most relevant traffic features and perform accurate traffic recognition makes our deep learning based approach an efficient, flexible and robust technique for website fingerprinting.
The influence of Deep Learning on image identification and natural language processing has attracted enormous attention globally. The convolution neural network that can learn without prior extraction of features fits well in response to the rapid iteration of Android malware. The traditional solution for detecting Android malware requires continuous learning through pre-extracted features to maintain high performance of identifying the malware. In order to reduce the manpower of feature engineering prior to the condition of not to extract pre-selected features, we have developed a coloR-inspired convolutional neuRal networks (CNN)-based AndroiD malware Detection (R2-D2) system. The system can convert the bytecode of classes.dex from Android archive file to rgb color code and store it as a color image with fixed size. The color image is input to the convolutional neural network for automatic feature extraction and training. The data was collected from Jan. 2017 to Aug 2017. During the period of time, we have collected approximately 2 million of benign and malicious Android apps for our experiments with the help from our research partner Leopard Mobile Inc. Our experiment results demonstrate that the proposed system has accurate security analysis on contracts. Furthermore, we keep our research results and experiment materials on http://R2D2.TWMAN.ORG.
Seizure prediction has attracted a growing attention as one of the most challenging predictive data analysis efforts in order to improve the life of patients living with drug-resistant epilepsy and tonic seizures. Many outstanding works have been reporting great results in providing a sensible indirect (warning systems) or direct (interactive neural-stimulation) control over refractory seizures, some of which achieved high performance. However, many works put heavily handcraft feature extraction and/or carefully tailored feature engineering to each patient to achieve very high sensitivity and low false prediction rate for a particular dataset. This limits the benefit of their approaches if a different dataset is used. In this paper we apply Convolutional Neural Networks (CNNs) on different intracranial and scalp electroencephalogram (EEG) datasets and proposed a generalized retrospective and patient-specific seizure prediction method. We use Short-Time Fourier Transform (STFT) on 30-second EEG windows with 50% overlapping to extract information in both frequency and time domains. A standardization step is then applied on STFT components across the whole frequency range to prevent high frequencies features being influenced by those at lower frequencies. A convolutional neural network model is used for both feature extraction and classification to separate preictal segments from interictal ones. The proposed approach achieves sensitivity of 81.4%, 81.2%, 82.3% and false prediction rate (FPR) of 0.06/h, 0.16/h, 0.22/h on Freiburg Hospital intracranial EEG (iEEG) dataset, Children's Hospital of Boston-MIT scalp EEG (sEEG) dataset, and Kaggle American Epilepsy Society Seizure Prediction Challenge's dataset, respectively. Our prediction method is also statistically better than an unspecific random predictor for most of patients in all three datasets.
Identifying the veracity of a news article is an interesting problem while automating this process can be a challenging task. Detection of a news article as fake is still an open question as it is contingent on many factors which the current state-of-the-art models fail to incorporate. In this paper, we explore a subtask to fake news identification, and that is stance detection. Given a news article, the task is to determine the relevance of the body and its claim. We present a novel idea that combines the neural, statistical and external features to provide an efficient solution to this problem. We compute the neural embedding from the deep recurrent model, statistical features from the weighted n-gram bag-of-words model and handcrafted external features with the help of feature engineering heuristics. Finally, using deep neural layer all the features are combined, thereby classifying the headline-body news pair as agree, disagree, discuss, or unrelated. We compare our proposed technique with the current state-of-the-art models on the fake news challenge dataset. Through extensive experiments, we find that the proposed model outperforms all the state-of-the-art techniques including the submissions to the fake news challenge.
Chemical-chemical interaction (CCI) plays a key role in predicting candidate drugs, toxicity, therapeutic effects, and biological functions. In various types of chemical analyses, computational approaches are often required due to the amount of data that needs to be handled. The recent remarkable growth and outstanding performance of deep learning have attracted considerable research attention. However,even in state-of-the-art drug analysis methods, deep learning continues to be used only as a classifier, although deep learning is capable of not only simple classification but also automated feature extraction. In this paper, we propose the first end-to-end learning method for CCI, named DeepCCI. Hidden features are derived from a simplified molecular input line entry system (SMILES), which is a string notation representing the chemical structure, instead of learning from crafted features. To discover hidden representations for the SMILES strings, we use convolutional neural networks (CNNs). To guarantee the commutative property for homogeneous interaction, we apply model sharing and hidden representation merging techniques. The performance of DeepCCI was compared with a plain deep classifier and conventional machine learning methods. The proposed DeepCCI showed the best performance in all seven evaluation metrics used. In addition, the commutative property was experimentally validated. The automatically extracted features through end-to-end SMILES learning alleviates the significant efforts required for manual feature engineering. It is expected to improve prediction performance, in drug analyses.
To assure cyber security of an enterprise, typically SIEM (Security Information and Event Management) system is in place to normalize security event from different preventive technologies and flag alerts. Analysts in the security operation center (SOC) investigate the alerts to decide if it is truly malicious or not. However, generally the number of alerts is overwhelming with majority of them being false positive and exceeding the SOC's capacity to handle all alerts. There is a great need to reduce the false positive rate as much as possible. While most previous research focused on network intrusion detection, we focus on risk detection and propose an intelligent Deep Belief Network machine learning system. The system leverages alert information, various security logs and analysts' investigation results in a real enterprise environment to flag hosts that have high likelihood of being compromised. Text mining and graph based method are used to generate targets and create features for machine learning. In the experiment, Deep Belief Network is compared with other machine learning algorithms, including multi-layer neural network, random forest, support vector machine and logistic regression. Results on real enterprise data indicate that the deep belief network machine learning system performs better than other algorithms for our problem and is six times more effective than current rule-based system. We also implement the whole system from data collection, label creation, feature engineering to host score generation in a real enterprise production environment.
In supervised approaches for keyphrase extraction, a candidate phrase is encoded with a set of hand-crafted features and machine learning algorithms are trained to discriminate keyphrases from non-keyphrases. Although the manually-designed features have shown to work well in practice, feature engineering is a difficult process that requires expert knowledge and normally does not generalize well. In this paper, we present SurfKE, a feature learning framework that exploits the text itself to automatically discover patterns that keyphrases exhibit. Our model represents the document as a graph and automatically learns feature representation of phrases. The proposed model obtains remarkable improvements in performance over strong baselines.
As an alternative to question answering methods based on feature engineering, deep learning approaches such as convolutional neural networks (CNNs) and Long Short-Term Memory Models (LSTMs) have recently been proposed for semantic matching of questions and answers. To achieve good results, however, these models have been combined with additional features such as word overlap or BM25 scores. Without this combination, these models perform significantly worse than methods based on linguistic feature engineering. In this paper, we propose an attention based neural matching model for ranking short answer text. We adopt value-shared weighting scheme instead of position-shared weighting scheme for combining different matching signals and incorporate question term importance learning using question attention network. Using the popular benchmark TREC QA data, we show that the relatively simple aNMM model can significantly outperform other neural network models that have been used for the question answering task, and is competitive with models that are combined with additional features. When aNMM is combined with additional features, it outperforms all baselines.
This paper presents HeNet, a hierarchical ensemble neural network, applied to classify hardware-generated control flow traces for malware detection. Deep learning-based malware detection has so far focused on analyzing executable files and runtime API calls. Static code analysis approaches face challenges due to obfuscated code and adversarial perturbations. Behavioral data collected during execution is more difficult to obfuscate but recent research has shown successful attacks against API call based malware classifiers. We investigate control flow based characterization of a program execution to build robust deep learning malware classifiers. HeNet consists of a low-level behavior model and a top-level ensemble model. The low-level model is a per-application behavior model, trained via transfer learning on a time-series of images generated from control flow trace of an execution. We use Intel$^\circledR$ Processor Trace enabled processor for low overhead execution tracing and design a lightweight image conversion and segmentation of the control flow trace. The top-level ensemble model aggregates the behavior classification of all the trace segments and detects an attack. The use of hardware trace adds portability to our system and the use of deep learning eliminates the manual effort of feature engineering. We evaluate HeNet against real-world exploitations of PDF readers. HeNet achieves 100\% accuracy and 0\% false positive on test set, and higher classification accuracy compared to classical machine learning algorithms.