PyTorchKR Papers

최신 논문

총 605편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1607.03428

Learning in Quantum Control: High-Dimensional Global Optimization for Noisy Quantum Dynamics

Quantum control is valuable for various quantum technologies such as high-fidelity gates for universal quantum computing, adaptive quantum-enhanced metrology, and ultra-cold atom manipulation. Although supervised machine learning and reinforcement learning are widely used for optimizing control parameters in classical systems, quantum control for parameter optimization is mainly pursued via gradient-based greedy algorithms. Although the quantum fitness landscape is often compatible with greedy algorithms, sometimes greedy algorithms yield poor results, especially for large-dimensional quantum systems. We employ differential evolution algorithms to circumvent the stagnation problem of non-convex optimization. We improve quantum control fidelity for noisy system by averaging over the objective function. To reduce computational cost, we introduce heuristics for early termination of runs and for adaptive selection of search subspaces. Our implementation is massively parallel and vectorized to reduce run time even further. We demonstrate our methods with two examples, namely quantum phase estimation and quantum gate design, for which we achieve superior fidelity and scalability than obtained using greedy algorithms.

arXiv:1611.09956

Efficient Likelihood Bayesian Constrained Local Model

The constrained local model (CLM) proposes a paradigm that the locations of a set of local landmark detectors are constrained to lie in a subspace, spanned by a shape point distribution model (PDM). Fitting the model to an object involves two steps. A response map, which represents the likelihood of the location of a landmark, is first computed for each landmark using local-texture detectors. Then, an optimal PDM is determined by jointly maximizing all the response maps simultaneously, with a global shape constraint. This global optimization can be considered as a Bayesian inference problem, where the posterior distribution of the shape parameters, as well as the pose parameters, can be inferred using maximum a posteriori (MAP). In this paper, we present a cascaded face-alignment approach, which employs random-forest regressors to estimate the positions of each landmark, as a likelihood term, efficiently in the CLM model. Interpretation from CLM framework, this algorithm is named as an efficient likelihood Bayesian constrained local model (elBCLM). Furthermore, in each stage of the regressors, the PDM non-rigid parameters of previous stage can work as shape clues for training each stage regressors. Experimental results on benchmarks show our approach achieve about 3 to 5 times speed-up compared with CLM models and improve around 10% on fitting quality compare with the same setting regression models.

arXiv:1612.03117

Advancing Bayesian Optimization: The Mixed-Global-Local (MGL) Kernel and Length-Scale Cool Down

Bayesian Optimization (BO) has become a core method for solving expensive black-box optimization problems. While much research focussed on the choice of the acquisition function, we focus on online length-scale adaption and the choice of kernel function. Instead of choosing hyperparameters in view of maximum likelihood on past data, we propose to use the acquisition function to decide on hyperparameter adaptation more robustly and in view of the future optimization progress. Further, we propose a particular kernel function that includes non-stationarity and local anisotropy and thereby implicitly integrates the efficiency of local convex optimization with global Bayesian optimization. Comparisons to state-of-the art BO methods underline the efficiency of these mechanisms on global optimization benchmarks.

arXiv:1607.04917

Piecewise convexity of artificial neural networks

Although artificial neural networks have shown great promise in applications including computer vision and speech recognition, there remains considerable practical and theoretical difficulty in optimizing their parameters. The seemingly unreasonable success of gradient descent methods in minimizing these non-convex functions remains poorly understood. In this work we offer some theoretical guarantees for networks with piecewise affine activation functions, which have in recent years become the norm. We prove three main results. Firstly, that the network is piecewise convex as a function of the input data. Secondly, that the network, considered as a function of the parameters in a single layer, all others held constant, is again piecewise convex. Finally, that the network as a function of all its parameters is piecewise multi-convex, a generalization of biconvexity. From here we characterize the local minima and stationary points of the training objective, showing that they minimize certain subsets of the parameter space. We then analyze the performance of two optimization algorithms on multi-convex problems: gradient descent, and a method which repeatedly solves a number of convex sub-problems. We prove necessary convergence conditions for the first algorithm and both necessary and sufficient conditions for the second, after introducing regularization to the objective. Finally, we remark on the remaining difficulty of the global optimization problem. Under the squared error objective, we show that by varying the training data, a single rectifier neuron admits local minima arbitrarily far apart, both in objective value and parameter space.

arXiv:1703.00329

Convergence rate of a simulated annealing algorithm with noisy observations

In this paper we propose a modified version of the simulated annealing algorithm for solving a stochastic global optimization problem. More precisely, we address the problem of finding a global minimizer of a function with noisy evaluations. We provide a rate of convergence and its optimized parametrization to ensure a minimal number of evaluations for a given accuracy and a confidence level close to 1. This work is completed with a set of numerical experimentations and assesses the practical performance both on benchmark test cases and on real world examples.

arXiv:1702.07360

Neural Decision Trees

In this paper we propose a synergistic melting of neural networks and decision trees (DT) we call neural decision trees (NDT). NDT is an architecture a la decision tree where each splitting node is an independent multilayer perceptron allowing oblique decision functions or arbritrary nonlinear decision function if more than one layer is used. This way, each MLP can be seen as a node of the tree. We then show that with the weight sharing asumption among those units, we end up with a Hashing Neural Network (HNN) which is a multilayer perceptron with sigmoid activation function for the last layer as opposed to the standard softmax. The output units then jointly represent the probability to be in a particular region. The proposed framework allows for global optimization as opposed to greedy in DT and differentiability w.r.t. all parameters and the input, allowing easy integration in any learnable pipeline, for example after CNNs for computer vision tasks. We also demonstrate the modeling power of HNN allowing to learn union of disjoint regions for final clustering or classification making it more general and powerful than standard softmax MLP requiring linear separability thus reducing the need on the inner layer to perform complex data transformations. We finally show experiments for supervised, semi-suppervised and unsupervised tasks and compare results with standard DTs and MLPs.

arXiv:1603.04381

A ranking approach to global optimization

We consider the problem of maximizing an unknown function over a compact and convex set using as few observations as possible. We observe that the optimization of the function essentially relies on learning the induced bipartite ranking rule of f. Based on this idea, we relate global optimization to bipartite ranking which allows to address problems with high dimensional input space, as well as cases of functions with weak regularity properties. The paper introduces novel meta-algorithms for global optimization which rely on the choice of any bipartite ranking method. Theoretical properties are provided as well as convergence guarantees and equivalences between various optimization methods are obtained as a by-product. Eventually, numerical evidence is given to show that the main algorithm of the paper which adapts empirically to the underlying ranking structure essentially outperforms existing state-of-the-art global optimization algorithms in typical benchmarks.

arXiv:1703.03038

Parallel Implementation of Efficient Search Schemes for the Inference of Cancer Progression Models

The emergence and development of cancer is a consequence of the accumulation over time of genomic mutations involving a specific set of genes, which provides the cancer clones with a functional selective advantage. In this work, we model the order of accumulation of such mutations during the progression, which eventually leads to the disease, by means of probabilistic graphic models, i.e., Bayesian Networks (BNs). We investigate how to perform the task of learning the structure of such BNs, according to experimental evidence, adopting a global optimization meta-heuristics. In particular, in this work we rely on Genetic Algorithms, and to strongly reduce the execution time of the inference -- which can also involve multiple repetitions to collect statistically significant assessments of the data -- we distribute the calculations using both multi-threading and a multi-node architecture. The results show that our approach is characterized by good accuracy and specificity; we also demonstrate its feasibility, thanks to a 84x reduction of the overall execution time with respect to a traditional sequential implementation.

arXiv:1703.04561

Drone Squadron Optimization: a Self-adaptive Algorithm for Global Numerical Optimization

This paper proposes Drone Squadron Optimization, a new self-adaptive metaheuristic for global numerical optimization which is updated online by a hyper-heuristic. DSO is an artifact-inspired technique, as opposed to many algorithms used nowadays, which are nature-inspired. DSO is very flexible because it is not related to behaviors or natural phenomena. DSO has two core parts: the semi-autonomous drones that fly over a landscape to explore, and the Command Center that processes the retrieved data and updates the drones' firmware whenever necessary. The self-adaptive aspect of DSO in this work is the perturbation/movement scheme, which is the procedure used to generate target coordinates. This procedure is evolved by the Command Center during the global optimization process in order to adapt DSO to the search landscape. DSO was evaluated on a set of widely employed benchmark functions. The statistical analysis of the results shows that the proposed method is competitive with the other methods in the comparison, the performance is promising, but several future improvements are planned.

arXiv:1703.07169

A Deterministic Global Optimization Method for Variational Inference

Variational inference methods for latent variable statistical models have gained popularity because they are relatively fast, can handle large data sets, and have deterministic convergence guarantees. However, in practice it is unclear whether the fixed point identified by the variational inference algorithm is a local or a global optimum. Here, we propose a method for constructing iterative optimization algorithms for variational inference problems that are guaranteed to converge to the $\epsilon$-global variational lower bound on the log-likelihood. We derive inference algorithms for two variational approximations to a standard Bayesian Gaussian mixture model (BGMM). We present a minimal data set for empirically testing convergence and show that a variational inference algorithm frequently converges to a local optimum while our algorithm always converges to the globally optimal variational lower bound. We characterize the loss incurred by choosing a non-optimal variational approximation distribution suggesting that selection of the approximating variational distribution deserves as much attention as the selection of the original statistical model for a given data set.

arXiv:1703.10764

A Hybrid Data Association Framework for Robust Online Multi-Object Tracking

Global optimization algorithms have shown impressive performance in data-association based multi-object tracking, but handling online data remains a difficult hurdle to overcome. In this paper, we present a hybrid data association framework with a min-cost multi-commodity network flow for robust online multi-object tracking. We build local target-specific models interleaved with global optimization of the optimal data association over multiple video frames. More specifically, in the min-cost multi-commodity network flow, the target-specific similarities are online learned to enforce the local consistency for reducing the complexity of the global data association. Meanwhile, the global data association taking multiple video frames into account alleviates irrecoverable errors caused by the local data association between adjacent frames. To ensure the efficiency of online tracking, we give an efficient near-optimal solution to the proposed min-cost multi-commodity flow problem, and provide the empirical proof of its sub-optimality. The comprehensive experiments on real data demonstrate the superior tracking performance of our approach in various challenging situations.

arXiv:1703.04842

Budgeted Batch Bayesian Optimization With Unknown Batch Sizes

Parameter settings profoundly impact the performance of machine learning algorithms and laboratory experiments. The classical grid search or trial-error methods are exponentially expensive in large parameter spaces, and Bayesian optimization (BO) offers an elegant alternative for global optimization of black box functions. In situations where the black box function can be evaluated at multiple points simultaneously, batch Bayesian optimization is used. Current batch BO approaches are restrictive in that they fix the number of evaluations per batch, and this can be wasteful when the number of specified evaluations is larger than the number of real maxima in the underlying acquisition function. We present the Budgeted Batch Bayesian Optimization (B3O) for hyper-parameter tuning and experimental design - we identify the appropriate batch size for each iteration in an elegant way. To set the batch size flexible, we use the infinite Gaussian mixture model (IGMM) for automatically identifying the number of peaks in the underlying acquisition functions. We solve the intractability of estimating the IGMM directly from the acquisition function by formulating the batch generalized slice sampling to efficiently draw samples from the acquisition function. We perform extensive experiments for both synthetic functions and two real world applications - machine learning hyper-parameter tuning and experimental design for alloy hardening. We show empirically that the proposed B3O outperforms the existing fixed batch BO approaches in finding the optimum whilst requiring a fewer number of evaluations, thus saving cost and time.

arXiv:1704.05318공식 코드 · Papers with Code

On the choice of the low-dimensional domain for global optimization via random embeddings

The challenge of taking many variables into account in optimization problems may be overcome under the hypothesis of low effective dimensionality. Then, the search of solutions can be reduced to the random embedding of a low dimensional space into the original one, resulting in a more manageable optimization problem. Specifically, in the case of time consuming black-box functions and when the budget of evaluations is severely limited, global optimization with random embeddings appears as a sound alternative to random search. Yet, in the case of box constraints on the native variables, defining suitable bounds on a low dimensional domain appears to be complex. Indeed, a small search domain does not guarantee to find a solution even under restrictive hypotheses about the function, while a larger one may slow down convergence dramatically. Here we tackle the issue of low-dimensional domain selection based on a detailed study of the properties of the random embedding, giving insight on the aforementioned difficulties. In particular, we describe a minimal low-dimensional set in correspondence with the embedded search space. We additionally show that an alternative equivalent embedding procedure yields simultaneously a simpler definition of the low-dimensional minimal set and better properties in practice. Finally, the performance and robustness gains of the proposed enhancements for Bayesian optimization are illustrated on numerical examples.

arXiv:1704.05724

The True Destination of EGO is Multi-local Optimization

Efficient global optimization is a popular algorithm for the optimization of expensive multimodal black-box functions. One important reason for its popularity is its theoretical foundation of global convergence. However, as the budgets in expensive optimization are very small, the asymptotic properties only play a minor role and the algorithm sometimes comes off badly in experimental comparisons. Many alternative variants have therefore been proposed over the years. In this work, we show experimentally that the algorithm instead has its strength in a setting where multiple optima are to be identified.

arXiv:1602.00853

An analytic comparison of regularization methods for Gaussian Processes

Gaussian Processes (GPs) are a popular approach to predict the output of a parameterized experiment. They have many applications in the field of Computer Experiments, in particular to perform sensitivity analysis, adaptive design of experiments and global optimization. Nearly all of the applications of GPs require the inversion of a covariance matrix that, in practice, is often ill-conditioned. Regularization methodologies are then employed with consequences on the GPs that need to be better understood.The two principal methods to deal with ill-conditioned covariance matrices are i) pseudoinverse and ii) adding a positive constant to the diagonal (the so-called nugget regularization).The first part of this paper provides an algebraic comparison of PI and nugget regularizations. Redundant points, responsible for covariance matrix singularity, are defined. It is proven that pseudoinverse regularization, contrarily to nugget regularization, averages the output values and makes the variance zero at redundant points. However, pseudoinverse and nugget regularizations become equivalent as the nugget value vanishes. A measure for data-model discrepancy is proposed which serves for choosing a regularization technique.In the second part of the paper, a distribution-wise GP is introduced that interpolates Gaussian distributions instead of data points. Distribution-wise GP can be seen as an improved regularization method for GPs.

arXiv:1705.07142

Simultaneous Multiple Surface Segmentation Using Deep Learning

The task of automatically segmenting 3-D surfaces representing boundaries of objects is important for quantitative analysis of volumetric images, and plays a vital role in biomedical image analysis. Recently, graph-based methods with a global optimization property have been developed and optimized for various medical imaging applications. Despite their widespread use, these require human experts to design transformations, image features, surface smoothness priors, and re-design for a different tissue, organ or imaging modality. Here, we propose a Deep Learning based approach for segmentation of the surfaces in volumetric medical images, by learning the essential features and transformations from training data, without any human expert intervention. We employ a regional approach to learn the local surface profiles. The proposed approach was evaluated on simultaneous intraretinal layer segmentation of optical coherence tomography (OCT) images of normal retinas and retinas affected by age related macular degeneration (AMD). The proposed approach was validated on 40 retina OCT volumes including 20 normal and 20 AMD subjects. The experiments showed statistically significant improvement in accuracy for our approach compared to state-of-the-art graph based optimal surface segmentation with convex priors (G-OSC). A single Convolution Neural Network (CNN) was used to learn the surfaces for both normal and diseased images. The mean unsigned surface positioning errors obtained by G-OSC method 2.31 voxels (95% CI 2.02-2.60 voxels) was improved to $1.27$ voxels (95% CI 1.14-1.40 voxels) using our new approach. On average, our approach takes 94.34 s, requiring 95.35 MB memory, which is much faster than the 2837.46 s and 6.87 GB memory required by the G-OSC method on the same computer system.

arXiv:1705.09549

Residual Expansion Algorithm: Fast and Effective Optimization for Nonconvex Least Squares Problems

We propose the residual expansion (RE) algorithm: a global (or near-global) optimization method for nonconvex least squares problems. Unlike most existing nonconvex optimization techniques, the RE algorithm is not based on either stochastic or multi-point searches; therefore, it can achieve fast global optimization. Moreover, the RE algorithm is easy to implement and successful in high-dimensional optimization. The RE algorithm exhibits excellent empirical performance in terms of k-means clustering, point-set registration, optimized product quantization, and blind image deblurring.

arXiv:1706.02246

Stochastic Global Optimization Algorithms: A Systematic Formal Approach

As we know, some global optimization problems cannot be solved using analytic methods, so numeric/algorithmic approaches are used to find near to the optimal solutions for them. A stochastic global optimization algorithm (SGoal) is an iterative algorithm that generates a new population (a set of candidate solutions) from a previous population using stochastic operations. Although some research works have formalized SGoals using Markov kernels, such formalization is not general and sometimes is blurred. In this paper, we propose a comprehensive and systematic formal approach for studying SGoals. First, we present the required theory of probability (\sigma-algebras, measurable functions, kernel, markov chain, products, convergence and so on) and prove that some algorithmic functions like swapping and projection can be represented by kernels. Then, we introduce the notion of join-kernel as a way of characterizing the combination of stochastic methods. Next, we define the optimization space, a formal structure (a set with a \sigma-algebra that contains strict \epsilon-optimal states) for studying SGoals, and we develop kernels, like sort and permutation, on such structure. Finally, we present some popular SGoals in terms of the developed theory, we introduce sufficient conditions for convergence of a SGoal, and we prove convergence of some popular SGoals.

arXiv:1611.03824

Learning to Learn without Gradient Descent by Gradient Descent

We learn recurrent neural network optimizers trained on simple synthetic functions by gradient descent. We show that these learned optimizers exhibit a remarkable degree of transfer in that they can be used to efficiently optimize a broad range of derivative-free black-box functions, including Gaussian process bandits, simple control objectives, global optimization benchmarks and hyper-parameter tuning tasks. Up to the training horizon, the learned optimizers learn to trade-off exploration and exploitation, and compare favourably with heavily engineered Bayesian optimization packages for hyper-parameter tuning.

arXiv:1610.09322

Homotopy Analysis for Tensor PCA

Developing efficient and guaranteed nonconvex algorithms has been an important challenge in modern machine learning. Algorithms with good empirical performance such as stochastic gradient descent often lack theoretical guarantees. In this paper, we analyze the class of homotopy or continuation methods for global optimization of nonconvex functions. These methods start from an objective function that is efficient to optimize (e.g. convex), and progressively modify it to obtain the required objective, and the solutions are passed along the homotopy path. For the challenging problem of tensor PCA, we prove global convergence of the homotopy method in the "high noise" regime. The signal-to-noise requirement for our algorithm is tight in the sense that it matches the recovery guarantee for the best degree-4 sum-of-squares algorithm. In addition, we prove a phase transition along the homotopy path for tensor PCA. This allows to simplify the homotopy method to a local search algorithm, viz., tensor power iterations, with a specific initialization and a noise injection procedure, while retaining the theoretical guarantees.