PyTorchKR Papers

최신 논문

총 605편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2305.11665

A Generic Performance Model for Deep Learning in a Distributed Environment

Performance modelling of a deep learning application is essential to improve and quantify the efficiency of the model framework. However, existing performance models are mostly case-specific, with limited capability for the new deep learning frameworks/applications. In this paper, we propose a generic performance model of an application in a distributed environment with a generic expression of the application execution time that considers the influence of both intrinsic factors/operations (e.g. algorithmic parameters/internal operations) and extrinsic scaling factors (e.g. the number of processors, data chunks and batch size). We formulate it as a global optimization problem and solve it using regularization on a cost function and differential evolution algorithm to find the best-fit values of the constants in the generic expression to match the experimentally determined computation time. We have evaluated the proposed model on three deep learning frameworks (i.e., TensorFlow, MXnet, and Pytorch). The experimental results show that the proposed model can provide accurate performance predictions and interpretability. In addition, the proposed work can be applied to any distributed deep neural network without instrumenting the code and provides insight into the factors affecting performance and scalability.

arXiv:2305.04557공식 코드 · Papers with Code

Toward Adversarial Training on Contextualized Language Representation

Beyond the success story of adversarial training (AT) in the recent text domain on top of pre-trained language models (PLMs), our empirical study showcases the inconsistent gains from AT on some tasks, e.g. commonsense reasoning, named entity recognition. This paper investigates AT from the perspective of the contextualized language representation outputted by PLM encoders. We find the current AT attacks lean to generate sub-optimal adversarial examples that can fool the decoder part but have a minor effect on the encoder. However, we find it necessary to effectively deviate the latter one to allow AT to gain. Based on the observation, we propose simple yet effective \textit{Contextualized representation-Adversarial Training} (CreAT), in which the attack is explicitly optimized to deviate the contextualized representation of the encoder. It allows a global optimization of adversarial examples that can fool the entire model. We also find CreAT gives rise to a better direction to optimize the adversarial examples, to let them less sensitive to hyperparameters. Compared to AT, CreAT produces consistent performance gains on a wider range of tasks and is proven to be more effective for language pre-training where only the encoder part is kept for downstream tasks. We achieve the new state-of-the-art performances on a series of challenging benchmarks, e.g. AdvGLUE (59.1 $ \rightarrow $ 61.1), HellaSWAG (93.0 $ \rightarrow $ 94.9), ANLI (68.1 $ \rightarrow $ 69.3).

arXiv:2305.04364

A Generalized Framework for Predictive Clustering and Optimization

Clustering is a powerful and extensively used data science tool. While clustering is generally thought of as an unsupervised learning technique, there are also supervised variations such as Spath's clusterwise regression that attempt to find clusters of data that yield low regression error on a supervised target. We believe that clusterwise regression is just a single vertex of a largely unexplored design space of supervised clustering models. In this article, we define a generalized optimization framework for predictive clustering that admits different cluster definitions (arbitrary point assignment, closest center, and bounding box) and both regression and classification objectives. We then present a joint optimization strategy that exploits mixed-integer linear programming (MILP) for global optimization in this generalized framework. To alleviate scalability concerns for large datasets, we also provide highly scalable greedy algorithms inspired by the Majorization-Minimization (MM) framework. Finally, we demonstrate the ability of our models to uncover different interpretable discrete cluster structures in data by experimenting with four real-world datasets.

arXiv:2305.03824공식 코드 · Papers with Code

No-Regret Constrained Bayesian Optimization of Noisy and Expensive Hybrid Models using Differentiable Quantile Function Approximations

This paper investigates the problem of efficient constrained global optimization of hybrid models that are a composition of a known white-box function and an expensive multi-output black-box function subject to noisy observations, which often arises in real-world science and engineering applications. We propose a novel method, Constrained Upper Quantile Bound (CUQB), to solve such problems that directly exploits the composite structure of the objective and constraint functions that we show leads substantially improved sampling efficiency. CUQB is a conceptually simple, deterministic approach that avoid constraint approximations used by previous methods. Although the CUQB acquisition function is not available in closed form, we propose a novel differentiable sample average approximation that enables it to be efficiently maximized. We further derive bounds on the cumulative regret and constraint violation under a non-parametric Bayesian representation of the black-box function. Since these bounds depend sublinearly on the number of iterations under some regularity assumptions, we establis bounds on the convergence rate to the optimal solution of the original constrained problem. In contrast to most existing methods, CUQB further incorporates a simple infeasibility detection scheme, which we prove triggers in a finite number of iterations when the original problem is infeasible (with high probability given the Bayesian model). Numerical experiments on several test problems, including environmental model calibration and real-time optimization of a reactor system, show that CUQB significantly outperforms traditional Bayesian optimization in both constrained and unconstrained cases. Furthermore, compared to other state-of-the-art methods that exploit composite structure, CUQB achieves competitive empirical performance while also providing substantially improved theoretical guarantees.

arXiv:2305.02894공식 코드 · Papers with Code

FedCBO: Reaching Group Consensus in Clustered Federated Learning through Consensus-based Optimization

Federated learning is an important framework in modern machine learning that seeks to integrate the training of learning models from multiple users, each user having their own local data set, in a way that is sensitive to data privacy and to communication loss constraints. In clustered federated learning, one assumes an additional unknown group structure among users, and the goal is to train models that are useful for each group, rather than simply training a single global model for all users. In this paper, we propose a novel solution to the problem of clustered federated learning that is inspired by ideas in consensus-based optimization (CBO). Our new CBO-type method is based on a system of interacting particles that is oblivious to group memberships. Our model is motivated by rigorous mathematical reasoning, including a mean field analysis describing the large number of particles limit of our particle system, as well as convergence guarantees for the simultaneous global optimization of general non-convex objective functions (corresponding to the loss functions of each cluster of users) in the mean-field regime. Experimental results demonstrate the efficacy of our FedCBO algorithm compared to other state-of-the-art methods and help validate our methodological and theoretical work.

arXiv:2305.02614

High-Dimensional Bayesian Optimization via Semi-Supervised Learning with Optimized Unlabeled Data Sampling

We introduce a novel semi-supervised learning approach, named Teacher-Student Bayesian Optimization ($\texttt{TSBO}$), integrating the teacher-student paradigm into BO to minimize expensive labeled data queries for the first time. $\texttt{TSBO}$ incorporates a teacher model, an unlabeled data sampler, and a student model. The student is trained on unlabeled data locations generated by the sampler, with pseudo labels predicted by the teacher. The interplay between these three components implements a unique selective regularization to the teacher in the form of student feedback. This scheme enables the teacher to predict high-quality pseudo labels, enhancing the generalization of the GP surrogate model in the search space. To fully exploit $\texttt{TSBO}$, we propose two optimized unlabeled data samplers to construct effective student feedback that well aligns with the objective of Bayesian optimization. Furthermore, we quantify and leverage the uncertainty of the teacher-student model for the provision of reliable feedback to the teacher in the presence of risky pseudo-label predictions. $\texttt{TSBO}$ demonstrates significantly improved sample-efficiency in several global optimization tasks under tight labeled data budgets.

arXiv:2305.00438

META-SMGO-$Δ$: similarity as a prior in black-box optimization

When solving global optimization problems in practice, one often ends up repeatedly solving problems that are similar to each others. By providing a rigorous definition of similarity, in this work we propose to incorporate the META-learning rationale into SMGO-$\Delta$, a global optimization approach recently proposed in the literature, to exploit priors obtained from similar past experience to efficiently solve new (similar) problems. Through a benchmark numerical example we show the practical benefits of our META-extension of the baseline algorithm, while providing theoretical bounds on its performance.

arXiv:2304.13114공식 코드 · Papers with Code

BO-ICP: Initialization of Iterative Closest Point Based on Bayesian Optimization

Typical algorithms for point cloud registration such as Iterative Closest Point (ICP) require a favorable initial transform estimate between two point clouds in order to perform a successful registration. State-of-the-art methods for choosing this starting condition rely on stochastic sampling or global optimization techniques such as branch and bound. In this work, we present a new method based on Bayesian optimization for finding the critical initial ICP transform. We provide three different configurations for our method which highlights the versatility of the algorithm to both find rapid results and refine them in situations where more runtime is available such as offline map building. Experiments are run on popular data sets and we show that our approach outperforms state-of-the-art methods when given similar computation time. Furthermore, it is compatible with other improvements to ICP, as it focuses solely on the selection of an initial transform, a starting point for all ICP-based methods.

arXiv:2304.08913

Computational and Exploratory Landscape Analysis of the GKLS Generator

The GKLS generator is one of the most used testbeds for benchmarking global optimization algorithms. In this paper, we conduct both a computational analysis and the Exploratory Landscape Analysis (ELA) of the GKLS generator. We utilize both canonically used and newly generated classes of GKLS-generated problems and show their use in benchmarking three state-of-the-art methods (from evolutionary and deterministic communities) in dimensions 5 and 10. We show that the GKLS generator produces ``needle in a haystack'' type problems that become extremely difficult to optimize in higher dimensions. Furthermore, we conduct the ELA on the GKLS generator and then compare it to the ELA of two other widely used benchmark sets (BBOB and CEC 2014), and discuss the meaningfulness of the results.

arXiv:2304.04751

DeepHive: A multi-agent reinforcement learning approach for automated discovery of swarm-based optimization policies

We present an approach for designing swarm-based optimizers for the global optimization of expensive black-box functions. In the proposed approach, the problem of finding efficient optimizers is framed as a reinforcement learning problem, where the goal is to find optimization policies that require a few function evaluations to converge to the global optimum. The state of each agent within the swarm is defined as its current position and function value within a design space and the agents learn to take favorable actions that maximize reward, which is based on the final value of the objective function. The proposed approach is tested on various benchmark optimization functions and compared to the performance of other global optimization strategies. Furthermore, the effect of changing the number of agents, as well as the generalization capabilities of the trained agents are investigated. The results show superior performance compared to the other optimizers, desired scaling when the number of agents is varied, and acceptable performance even when applied to unseen functions. On a broader scale, the results show promise for the rapid development of domain-specific optimizers.

arXiv:2304.05346

Leo: Lagrange Elementary Optimization

Global optimization problems are frequently solved using the practical and efficient method of evolutionary sophistication. But as the original problem becomes more complex, so does its efficacy and expandability. Thus, the purpose of this research is to introduce the Lagrange Elementary Optimization (Leo) as an evolutionary method, which is self-adaptive inspired by the remarkable accuracy of vaccinations using the albumin quotient of human blood. They develop intelligent agents using their fitness function value after gene crossing. These genes direct the search agents during both exploration and exploitation. The main objective of the Leo algorithm is presented in this paper along with the inspiration and motivation for the concept. To demonstrate its precision, the proposed algorithm is validated against a variety of test functions, including 19 traditional benchmark functions and the CECC06 2019 test functions. The results of Leo for 19 classic benchmark test functions are evaluated against DA, PSO, and GA separately, and then two other recent algorithms such as FDO and LPB are also included in the evaluation. In addition, the Leo is tested by ten functions on CECC06 2019 with DA, WOA, SSA, FDO, LPB, and FOX algorithms distinctly. The cumulative outcomes demonstrate Leo's capacity to increase the starting population and move toward the global optimum. Different standard measurements are used to verify and prove the stability of Leo in both the exploration and exploitation phases. Moreover, Statistical analysis supports the findings results of the proposed research. Finally, novel applications in the real world are introduced to demonstrate the practicality of Leo.

arXiv:2303.14293

Efficient Lipschitzian Global Optimization of Hölder Continuous Multivariate Functions

This study presents an effective global optimization technique designed for multivariate functions that are H\"older continuous. Unlike traditional methods that construct lower bounding proxy functions, this algorithm employs a predetermined query creation rule that makes it computationally superior. The algorithm's performance is assessed using the average or cumulative regret, which also implies a bound for the simple regret and reflects the overall effectiveness of the approach. The results show that with appropriate parameters the algorithm attains an average regret bound of $O(T^{-\frac{\alpha}{n}})$ for optimizing a H\"older continuous target function with H\"older exponent $\alpha$ in an $n$-dimensional space within a given time horizon $T$. We demonstrate that this bound is minimax optimal.

arXiv:2303.14793

Multi-Phase Relaxation Labeling for Square Jigsaw Puzzle Solving

We present a novel method for solving square jigsaw puzzles based on global optimization. The method is fully automatic, assumes no prior information, and can handle puzzles with known or unknown piece orientation. At the core of the optimization process is nonlinear relaxation labeling, a well-founded approach for deducing global solutions from local constraints, but unlike the classical scheme here we propose a multi-phase approach that guarantees convergence to feasible puzzle solutions. Next to the algorithmic novelty, we also present a new compatibility function for the quantification of the affinity between adjacent puzzle pieces. Competitive results and the advantage of the multi-phase approach are demonstrated on standard datasets.

arXiv:2303.15416공식 코드 · Papers with Code

3D Video Object Detection with Learnable Object-Centric Global Optimization

We explore long-term temporal visual correspondence-based optimization for 3D video object detection in this work. Visual correspondence refers to one-to-one mappings for pixels across multiple images. Correspondence-based optimization is the cornerstone for 3D scene reconstruction but is less studied in 3D video object detection, because moving objects violate multi-view geometry constraints and are treated as outliers during scene reconstruction. We address this issue by treating objects as first-class citizens during correspondence-based optimization. In this work, we propose BA-Det, an end-to-end optimizable object detector with object-centric temporal correspondence learning and featuremetric object bundle adjustment. Empirically, we verify the effectiveness and efficiency of BA-Det for multiple baseline 3D detectors under various setups. Our BA-Det achieves SOTA performance on the large-scale Waymo Open Dataset (WOD) with only marginal computation cost. Our code is available at https://github.com/jiaweihe1996/BA-Det.

arXiv:2303.10024

Counter-example guided inductive synthesis of control Lyapunov functions for uncertain systems

We propose a counter-example guided inductive synthesis (CEGIS) scheme for the design of control Lyapunov functions and associated state-feedback controllers for linear systems affected by parametric uncertainty with arbitrary shape. In the CEGIS framework, a learner iteratively proposes a candidate control Lyapunov function and a tailored controller by solving a linear matrix inequality (LMI) feasibility problem, while a verifier either falsifies the current candidate by producing a counter-example to be considered at the next iteration, or it certifies that the tentative control Lyapunov function actually enjoys such feature. We investigate the Lipschitz continuity of the objective function of the global optimization problem solved by the verifier, which is key to establish the convergence of our method in a finite number of iterations. Numerical simulations confirm the effectiveness of the proposed approach.

arXiv:2303.04623

Continuous Function Structured in Multilayer Perceptron for Global Optimization

The gradient information of multilayer perceptron with a linear neuron is modified with functional derivative for the global minimum search benchmarking problems. From this approach, we show that the landscape of the gradient derived from given continuous function using functional derivative can be the MLP-like form with ax+b neurons. In this extent, the suggested algorithm improves the availability of the optimization process to deal all the parameters in the problem set simultaneously. The functionality of this method could be improved through intentionally designed convex function with Kullack-Liebler divergence applied to cost value as well.

arXiv:2303.01682공식 코드 · Papers with Code

Neural-BO: A Black-box Optimization Algorithm using Deep Neural Networks

Bayesian Optimization (BO) is an effective approach for global optimization of black-box functions when function evaluations are expensive. Most prior works use Gaussian processes to model the black-box function, however, the use of kernels in Gaussian processes leads to two problems: first, the kernel-based methods scale poorly with the number of data points and second, kernel methods are usually not effective on complex structured high dimensional data due to curse of dimensionality. Therefore, we propose a novel black-box optimization algorithm where the black-box function is modeled using a neural network. Our algorithm does not need a Bayesian neural network to estimate predictive uncertainty and is therefore computationally favorable. We analyze the theoretical behavior of our algorithm in terms of regret bound using advances in NTK theory showing its efficient convergence. We perform experiments with both synthetic and real-world optimization tasks and show that our algorithm is more sample efficient compared to existing methods.

arXiv:2303.00179

A Unified Momentum-based Paradigm of Decentralized SGD for Non-Convex Models and Heterogeneous Data

Emerging distributed applications recently boosted the development of decentralized machine learning, especially in IoT and edge computing fields. In real-world scenarios, the common problems of non-convexity and data heterogeneity result in inefficiency, performance degradation, and development stagnation. The bulk of studies concentrates on one of the issues mentioned above without having a more general framework that has been proven optimal. To this end, we propose a unified paradigm called UMP, which comprises two algorithms, D-SUM and GT-DSUM, based on the momentum technique with decentralized stochastic gradient descent(SGD). The former provides a convergence guarantee for general non-convex objectives. At the same time, the latter is extended by introducing gradient tracking, which estimates the global optimization direction to mitigate data heterogeneity(i.e., distribution drift). We can cover most momentum-based variants based on the classical heavy ball or Nesterov's acceleration with different parameters in UMP. In theory, we rigorously provide the convergence analysis of these two approaches for non-convex objectives and conduct extensive experiments, demonstrating a significant improvement in model accuracy by up to 57.6% compared to other methods in practice.

arXiv:2302.13537

Global optimization in the discrete and variable-dimension conformational space: The case of crystal with the strongest atomic cohesion

We introduce a computational method to optimize target physical properties in the full configuration space regarding atomic composition, chemical stoichiometry, and crystal structure. The approach combines the universal potential of the crystal graph neural network and Bayesian optimization. The proposed approach effectively obtains the crystal structure with the strongest atomic cohesion from all possible crystals. Several new crystals with high atomic cohesion are identified and confirmed by density functional theory for thermodynamic and dynamic stability. Our method introduces a novel approach to inverse materials design with additional functional properties for practical applications.

arXiv:2302.11566공식 코드 · Papers with Code

Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via Self-supervised Scene Decomposition

We present Vid2Avatar, a method to learn human avatars from monocular in-the-wild videos. Reconstructing humans that move naturally from monocular in-the-wild videos is difficult. Solving it requires accurately separating humans from arbitrary backgrounds. Moreover, it requires reconstructing detailed 3D surface from short video sequences, making it even more challenging. Despite these challenges, our method does not require any groundtruth supervision or priors extracted from large datasets of clothed human scans, nor do we rely on any external segmentation modules. Instead, it solves the tasks of scene decomposition and surface reconstruction directly in 3D by modeling both the human and the background in the scene jointly, parameterized via two separate neural fields. Specifically, we define a temporally consistent human representation in canonical space and formulate a global optimization over the background model, the canonical human shape and texture, and per-frame human pose parameters. A coarse-to-fine sampling strategy for volume rendering and novel objectives are introduced for a clean separation of dynamic human and static background, yielding detailed and robust 3D human geometry reconstructions. We evaluate our methods on publicly available datasets and show improvements over prior art.