arXiv subjects

Phuong Nguyen

Publications and source records attributed to Phuong Nguyen.

32 records · Page 2Linked to original sources

arXiv2021

Tolerating Adversarial Attacks and Byzantine Faults in Distributed Machine Learning

Adversarial attacks attempt to disrupt the training, retraining and utilizing of artificial intelligence and machine learning models in large-scale distributed machine learning systems. This causes security risks on its prediction outcome. For example, attackers attempt to poison the model by either presenting inaccurate misrepresentative data or altering the models' parameters. In addition, Byzantine faults including software, hardware, network issues occur in distributed systems which also lead to a negative impact on the prediction outcome. In this paper, we propose a novel distributed training algorithm, partial synchronous stochastic gradient descent (ParSGD), which defends adversarial attacks and/or tolerates Byzantine faults. We demonstrate the effectiveness of our algorithm under three common adversarial attacks again the ML models and a Byzantine fault during the training phase. Our results show that using ParSGD, ML models can still produce accurate predictions as if it is not being attacked nor having failures at all when almost half of the nodes are being compromised or failed. We will report the experimental evaluations of ParSGD in comparison with other algorithms.

cs.CR↗

arXiv2021

Nonlinear Stochastic parabolic partial differential equations with a monotone operator of the Ladyzenskaya-Smagorinsky type, driven by a Levy noise

The aim of this article is to show the global existence of both martingale and pathwise solutions of stochastic equations with a monotone operator, of the Ladyzenskaya-Smagorinsky type, driven by a general Levy noise. The classical approach based on using directly the Galerkin approximation is not valid. Instead, our approach is based on using appropriate approximations for the monotone operator, Galerkin approximations and on the theory of martingale solutions.

math.AP↗

arXiv2021

Toward Generating Synthetic CT Volumes using a 3D-Conditional Generative Adversarial Network

We present a novel conditional Generative Adversarial Network (cGAN) architecture that is capable of generating 3D Computed Tomography scans in voxels from noisy and/or pixelated approximations and with the potential to generate full synthetic 3D scan volumes. We believe conditional cGAN to be a tractable approach to generate 3D CT volumes, even though the problem of generating full resolution deep fakes is presently impractical due to GPU memory limitations. We present results for autoencoder, denoising, and depixelating tasks which are trained and tested on two novel COVID19 CT datasets. Our evaluation metrics, Peak Signal to Noise ratio (PSNR) range from 12.53 - 46.46 dB, and the Structural Similarity index ( SSIM) range from 0.89 to 1.

eess.IV↗

arXiv2020

Lung Nodule Classification Using Biomarkers, Volumetric Radiomics and 3D CNNs

We present a hybrid algorithm to estimate lung nodule malignancy that combines imaging biomarkers from Radiologist's annotation with image classification of CT scans. Our algorithm employs a 3D Convolutional Neural Network (CNN) as well as a Random Forest in order to combine CT imagery with biomarker annotation and volumetric radiomic features. We analyze and compare the performance of the algorithm using only imagery, only biomarkers, combined imagery + biomarkers, combined imagery + volumetric radiomic features and finally the combination of imagery + biomarkers + volumetric features in order to classify the suspicion level of nodule malignancy. The National Cancer Institute (NCI) Lung Image Database Consortium (LIDC) IDRI dataset is used to train and evaluate the classification task. We show that the incorporation of semi-supervised learning by means of K-Nearest-Neighbors (KNN) can increase the available training sample size of the LIDC-IDRI thereby further improving the accuracy of malignancy estimation of most of the models tested although there is no significant improvement with the use of KNN semi-supervised learning if image classification with CNNs and volumetric features are combined with descriptive biomarkers. Unexpectedly, we also show that a model using image biomarkers alone is more accurate than one that combines biomarkers with volumetric radiomics, 3D CNNs, and semi-supervised learning. We discuss the possibility that this result may be influenced by cognitive bias in LIDC-IDRI because malignancy estimates were recorded by the same radiologist panel as biomarkers, as well as future work to incorporate pathology information over a subset of study participants.

eess.IV↗

arXiv2020

Deep Expectation-Maximization for Semi-Supervised Lung Cancer Screening

We present a semi-supervised algorithm for lung cancer screening in which a 3D Convolutional Neural Network (CNN) is trained using the Expectation-Maximization (EM) meta-algorithm. Semi-supervised learning allows a smaller labelled data-set to be combined with an unlabeled data-set in order to provide a larger and more diverse training sample. EM allows the algorithm to simultaneously calculate a maximum likelihood estimate of the CNN training coefficients along with the labels for the unlabeled training set which are defined as a latent variable space. We evaluate the model performance of the Semi-Supervised EM algorithm for CNNs through cross-domain training of the Kaggle Data Science Bowl 2017 (Kaggle17) data-set with the National Lung Screening Trial (NLST) data-set. Our results show that the Semi-Supervised EM algorithm greatly improves the classification accuracy of the cross-domain lung cancer screening, although results are lower than a fully supervised approach with the advantage of additional labelled data from the unsupervised sample. As such, we demonstrate that Semi-Supervised EM is a valuable technique to improve the accuracy of lung cancer screening models using 3D CNNs.

cs.LG↗

arXiv2020

Generating Realistic COVID19 X-rays with a Mean Teacher + Transfer Learning GAN

COVID-19 is a novel infectious disease responsible for over 800K deaths worldwide as of August 2020. The need for rapid testing is a high priority and alternative testing strategies including X-ray image classification are a promising area of research. However, at present, public datasets for COVID19 x-ray images have low data volumes, making it challenging to develop accurate image classifiers. Several recent papers have made use of Generative Adversarial Networks (GANs) in order to increase the training data volumes. But realistic synthetic COVID19 X-rays remain challenging to generate. We present a novel Mean Teacher + Transfer GAN (MTT-GAN) that generates COVID19 chest X-ray images of high quality. In order to create a more accurate GAN, we employ transfer learning from the Kaggle Pneumonia X-Ray dataset, a highly relevant data source orders of magnitude larger than public COVID19 datasets. Furthermore, we employ the Mean Teacher algorithm as a constraint to improve stability of training. Our qualitative analysis shows that the MTT-GAN generates X-ray images that are greatly superior to a baseline GAN and visually comparable to real X-rays. Although board-certified radiologists can distinguish MTT-GAN fakes from real COVID19 X-rays. Quantitative analysis shows that MTT-GAN greatly improves the accuracy of both a binary COVID19 classifier as well as a multi-class Pneumonia classifier as compared to a baseline GAN. Our classification accuracy is favourable as compared to recently reported results in the literature for similar binary and multi-class COVID19 screening tasks.

cs.LG↗

arXiv2019

SUMMARIZED: Efficient Framework for Analyzing Multidimensional Process Traces under Edit-distance Constraint

Domains such as scientific workflows and business processes exhibit data models with complex relationships between objects. This relationship is typically represented as sequences, where each data item is annotated with multi-dimensional attributes. There is a need to analyze this data for operational insights. For example, in business processes, users are interested in clustering process traces into smaller subsets to discover less complex process models. This requires expensive computation of similarity metrics between sequence-based data. Related work on dimension reduction and embedding methods do not take into account the multi-dimensional attributes of data, and do not address the interpretability of data in the embedding space (i.e., by favoring vector-based representation). In this work, we introduce Summarized, a framework for efficient analysis on sequence-based multi-dimensional data using intuitive and user-controlled summarizations. We introduce summarization schemes that provide tunable trade-offs between the quality and efficiency of analysis tasks and derive an error model for summary-based similarity under an edit-distance constraint. Evaluations using real-world datasets show the effectives of our framework.

cs.DB↗

arXiv2017

DAC-h3: A Proactive Robot Cognitive Architecture to Acquire and Express Knowledge About the World and the Self

This paper introduces a cognitive architecture for a humanoid robot to engage in a proactive, mixed-initiative exploration and manipulation of its environment, where the initiative can originate from both the human and the robot. The framework, based on a biologically-grounded theory of the brain and mind, integrates a reactive interaction engine, a number of state-of-the-art perceptual and motor learning algorithms, as well as planning abilities and an autobiographical memory. The architecture as a whole drives the robot behavior to solve the symbol grounding problem, acquire language capabilities, execute goal-oriented behavior, and express a verbal narrative of its own experience in the world. We validate our approach in human-robot interaction experiments with the iCub humanoid robot, showing that the proposed cognitive architecture can be applied in real time within a realistic scenario and that it can be used with naive users.

cs.AI↗

arXiv2017

Crowdsensing in Opportunistic Mobile Social Networks: A Context-aware and Human-centric Approach

In recent years, there have been efforts to collect human contact traces during social events (e.g., conferences) using Bluetooth devices (e.g., mobile phones, iMotes). The results of these studies have enabled the ability to do the crowd-sourcing task from within the crowd, in order to answer questions, such as: what is the current density of the crowd, or how many people are attending the event? However, in those studies, the sensing devices are usually distributed and configured in a certain manner. For example, the number of devices is fixed, people register for the devices on a volunteering basis. In this paper, we treat the above problem as an optimization problem and draw the connection to the vertex cover problem in graph theory. Since finding the optimal solution for minimum vertex cover problem is NP-complete, approximation algorithms have to be used. However, we will show that the well-known approximation algorithms do not perform well with the crowd-sensing task. In this paper, we propose the notions of node observability and coverage utility score and design a new context-aware approximation algorithm to find vertex cover that is tailored for crowd-sensing task. In addition, we design human-centric bootstrapping strategies to make initial assignment of sensing devices based on meta information about the participants (e.g., interests, friendship). The motivation is to assign the sensing task to a more "socialized" device to obtain better sensing coverage. We perform comprehensive experiments on real-world data traces obtained from previous experimental studies in conference and academic social context. The results show that our proposed approach significantly outperforms the baseline approximation algorithms in terms of sensing coverage.

cs.SI↗

arXiv2013

Optimal Regret Bounds for Selecting the State Representation in Reinforcement Learning

We consider an agent interacting with an environment in a single stream of actions, observations, and rewards, with no reset. This process is not assumed to be a Markov Decision Process (MDP). Rather, the agent has several representations (mapping histories of past interactions to a discrete state space) of the environment with unknown dynamics, only some of which result in an MDP. The goal is to minimize the average regret criterion against an agent who knows an MDP representation giving the highest optimal reward, and acts optimally in it. Recent regret bounds for this setting are of order $O(T^{2/3})$ with an additive term constant yet exponential in some characteristics of the optimal MDP. We propose an algorithm whose regret after $T$ time steps is $O(\sqrt{T})$, with all constants reasonably small. This is optimal in $T$ since $O(\sqrt{T})$ is the optimal regret in the setting of learning in a (single discrete) MDP.

cs.LG↗

arXiv2012

Relativizing Small Complexity Classes and their Theories

Existing definitions of the relativizations of \NCOne, Ł and \NL\ do not preserve the inclusions $\NCOne \subseteq Ł$, $\NL\subseteq \ACOne$. We start by giving the first definitions that preserve them. Here for Ł and \NL\ we define their relativizations using Wilson's stack oracle model, but limit the height of the stack to a constant (instead of $\log(n)$). We show that the collapse of any two classes in $\{\ACZm, \TCZ, \NCOne, Ł, \NL\}$ implies the collapse of their relativizations. Next we exhibit an oracle $α$ that makes $\ACk(α)$ a proper hierarchy. This strengthens and clarifies the separations of the relativized theories in [Takeuti, 1995]. The idea is that a circuit whose nested depth of oracle gates is bounded by $k$ cannot compute correctly the $(k+1)$ compositions of every oracle function. Finally we develop theories that characterize the relativizations of subclasses of \Ptime\ by modifying theories previously defined by the second two authors. A function is provably total in a theory iff it is in the corresponding relativized class, and hence the oracle separations imply separations for the relativized theories.

cs.CC↗

arXiv2011

Feature Reinforcement Learning In Practice

Following a recent surge in using history-based methods for resolving perceptual aliasing in reinforcement learning, we introduce an algorithm based on the feature reinforcement learning framework called PhiMDP. To create a practical algorithm we devise a stochastic search procedure for a class of context trees based on parallel tempering and a specialized proposal distribution. We provide the first empirical evaluation for PhiMDP. Our proposed algorithm achieves superior performance to the classical U-tree algorithm and the recent active-LZ algorithm, and is competitive with MC-AIXI-CTW that maintains a bayesian mixture over all context trees up to a chosen depth.We are encouraged by our ability to compete with this sophisticated method using an algorithm that simply picks one single model, and uses Q-learning on the corresponding MDP. Our PhiMDP algorithm is much simpler, yet consumes less time and memory. These results show promise for our future work on attacking more complex and larger problems.

cs.AI↗

arXiv2010

The Complexity of Proving the Discrete Jordan Curve Theorem

The Jordan Curve Theorem (JCT) states that a simple closed curve divides the plane into exactly two connected regions. We formalize and prove the theorem in the context of grid graphs, under different input settings, in theories of bounded arithmetic that correspond to small complexity classes. The theory $V^0(2)$ (corresponding to $AC^0(2)$) proves that any set of edges that form disjoint cycles divides the grid into at least two regions. The theory $V^0$ (corresponding to $AC^0$) proves that any sequence of edges that form a simple closed curve divides the grid into exactly two regions. As a consequence, the Hex tautologies and the st-connectivity tautologies have polynomial size $AC^0(2)$-Frege-proofs, which improves results of Buss which only apply to the stronger proof system $TC^0$-Frege.

cs.LO↗

arXiv2006

Theories for TC0 and Other Small Complexity Classes

We present a general method for introducing finitely axiomatizable "minimal" two-sorted theories for various subclasses of P (problems solvable in polynomial time). The two sorts are natural numbers and finite sets of natural numbers. The latter are essentially the finite binary strings, which provide a natural domain for defining the functions and sets in small complexity classes. We concentrate on the complexity class TC^0, whose problems are defined by uniform polynomial-size families of bounded-depth Boolean circuits with majority gates. We present an elegant theory VTC^0 in which the provably-total functions are those associated with TC^0, and then prove that VTC^0 is "isomorphic" to a different-looking single-sorted theory introduced by Johannsen and Pollet. The most technical part of the isomorphism proof is defining binary number multiplication in terms a bit-counting function, and showing how to formalize the proofs of its algebraic properties.

cs.LO↗