检索结果-内蒙古大学图书馆

Informatik-Spektrum 2020年第2期43卷 137-144页

作者： Humm, Bernhard Bense, Hermann Bock, Jürgen Classen, Mario Halvani, Oren Herta, Christian Hoppe, Thomas Juwig, Oliver Siegel, Melanie Hochschule Darmstadt—University of Applied Sciences Haardtring 100 Darmstadt64295 Germany *** GmbH Schwarze-Brüder-Str. 1 Dortmund44137 Germany KUKA Deutschland GmbH Augsburg Germany AXA Konzern AG Colonia Allee 10–20 Cologne51067 Germany Fraunhofer Institut für Sichere Informationstechnologie SIT Darmstadt Germany Hochschule für Technik und Wirtschaft Berlin Wilhelminenhofstr. 75a Berlin12459 Germany Fraunhofer FOKUS Kaiserin-Augusta-Allee 31 Berlin10589 Germany Hochschule für Technik und Wirtschaft Berlin Wilhelminenhofstr. 75a Berlin12459 Germany AXA Konzern AG Colonia Allee 10–20 Cologne51067 Germany

The relevance of Machine Intelligence, a.k.a. Artificial Intelligence (AI), is undisputed at the present time. This is not only due to AI successes in research but, more prominently, its use in day-to-day practice. In 2014, we started a series of annual workshops at the Leibniz Zentrum für Informatik, Schloss Dagstuhl, Germany, initially focussing on Corporate Semantic Web and later widening the scope to Applied Machine Intelligence. This article presents a number of AI applications from various application domains, including medicine, industrial manufacturing and the insurance sector. Best practices, current trends, possibilities and limitations of new AI approaches for developing AI applications are also presented. Focus is put on the areas of natural language processing, ontologies and machine learning. The article concludes with a summary and outlook. © 2020, The Author(s).

关键词： learning algorithms

Response outcomes gate the impact of expectations on perceptual decisions

学校读者我要写书评

暂无评论

NATURE COMMUNICATIONS 2020年第1期11卷 1057页

作者： Hermoso-Mendizabal, Ainhoa Hyafil, Alexandre Rueda-Orozco, Pavel E. Jaramillo, Santiago Robbe, David de la Rocha, Jaime Inst Invest Biomed August Pi i Sunyer IDIBAP Barcelona 08036 Spain Univ Pompeu Fabra Ctr Brain & Cognit Ramon Trias Fargas 25 Barcelona 08018 Spain UNAM Inst Neurobiol Santiago De Queretaro Me 76230 Mexico Univ Oregon Inst Neurosci 1254 Univ Oregon Eugene OR 97403 USA Aix Marseille Univ INMED INSERM 63 Ave Luminy F-13009 Marseille France Ctr Recerca Matemat Campus Bellaterra Bellaterra 08193 Spain

Perceptual decisions are based on sensory information but can also be influenced by expectations built from recent experiences. Can the impact of expectations be flexibly modulated based on the outcome of previous decisions? Here, rats perform an auditory task where the probability to repeat the previous stimulus category is varied in trial-blocks. All rats capitalize on these sequence correlations by exploiting a transition bias: a tendency to repeat or alternate their previous response using an internal estimate of the sequence repeating probability. Surprisingly, this bias is null after error trials. The internal estimate however is not reset and it becomes effective again after the next correct response. This behavior is captured by a generative model, whereby a reward-driven modulatory signal gates the impact of the latent model of the environment on the current decision. These results demonstrate that, based on previous outcomes, rats flexibly modulate how expectations influence their decisions.

关键词： Decision learning algorithms Perception

Revealing Secrets From Pre-trained Models

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Al Rafi, Mujahid Feng, Yuan Jeon, Hyeran University of California Merced United States

With the growing burden of training deep learning models with large data sets, transfer-learning has been widely adopted in many emerging deep learning algorithms. Transformer models such as BERT are the main player in natural language processing and use transfer-learning as a de facto standard training method. A few big data companies release pre-trained models that are trained with a few popular datasets with which end users and researchers fine-tune the model with their own datasets. Transfer-learning significantly reduces the time and effort of training models. However, it comes at the cost of security concerns. In this paper, we show a new observation that pre-trained models and fine-tuned models have significantly high similarities in weight values. Also, we demonstrate that there exist vendor-specific computing patterns even for the same models. With these new findings, we propose a new model extraction attack that reveals the model architecture and the pre-trained model used by the black-box victim model with vendor-specific computing patterns and then estimates the entire model weights based on the weight value similarities between the fine-tuned model and pre-trained model. We also show that the weight similarity can be leveraged for increasing the model extraction feasibility through a novel weight extraction pruning. © 2022, CC BY-NC-ND.

关键词： learning algorithms

A New Look at Dynamic Regret for Non-Stationary Stochastic Bandits

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Abbasi-Yadkori, Yasin György, András Lazić, Nevena DeepMind United Kingdom

We study the non-stationary stochastic multi-armed bandit problem, where the reward statistics of each arm may change several times during the course of learning. The performance of a learning algorithm is evaluated in terms of their dynamic regret, which is defined as the difference between the expected cumulative reward of an agent choosing the optimal arm in every time step and the cumulative reward of the learning algorithm. One way to measure the hardness of such environments is to consider how many times the identity of the optimal arm can change. We propose a method that achieves, in K-armed bandit problems, a near-optimal Oe(pKN(S + 1)) dynamic regret, where N is the time horizon of the problem and S is the number of times the identity of the optimal arm changes, without prior knowledge of S. Previous works for this problem obtain regret bounds that scale with the number of changes (or the amount of change) in the reward functions, which can be much larger, or assume prior knowledge of S to achieve similar bounds. Copyright © 2022, The Authors. All rights reserved.

关键词： learning algorithms

Adaptive Fine-Tuning of Transformer-Based Language Models for Named Entity Recognition

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Stollenwerk, Felix Arbetsförmedlingen Swedish Public Employment Service Sweden

The current standard approach for fine-tuning transformer-based language models includes a fixed number of training epochs and a linear learning rate schedule. In order to obtain a near-optimal model for the given downstream task, a search in optimization hyperparameter space is usually required. In particular, the number of training epochs needs to be adjusted to the dataset size. In this paper, we introduce adaptive fine-tuning, which is an alternative approach that uses early stopping and a custom learning rate schedule to dynamically adjust the number of training epochs to the dataset size. For the example use case of named entity recognition, we show that our approach not only makes hyperparameter search with respect to the number of training epochs redundant, but also leads to improved results in terms of performance, stability and efficiency. This holds true especially for small datasets, where we outperform the state-of-the-art fine-tuning method by a large margin. © 2022, CC BY.

关键词： learning algorithms

PyGlove: Symbolic programming for automated machine learning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Peng, Daiyi Dong, Xuanyi Real, Esteban Tan, Mingxing Lu, Yifeng Liu, Hanxiao Bender, Gabriel Kraft, Adam Liang, Chen Le, Quoc V. Google Research Brain Team United States

Neural networks are sensitive to hyper-parameter and architecture choices. Automated Machine learning (AutoML) is a promising paradigm for automating these choices. Current ML software libraries, however, are quite limited in handling the dynamic interactions among the components of AutoML. For example, efficient NAS algorithms, such as ENAS [1] and DARTS [2], typically require an implementation coupling between the search space and search algorithm, the two key components in AutoML. Furthermore, implementing a complex search flow, such as searching architectures within a loop of searching hardware configurations, is difficult. To summarize, changing the search space, search algorithm, or search flow in current ML libraries usually requires a significant change in the program logic. In this paper, we introduce a new way of programming AutoML based on symbolic programming. Under this paradigm, ML programs are mutable, thus can be manipulated easily by another program. As a result, AutoML can be reformulated as an automated process of symbolic manipulation. With this formulation, we decouple the triangle of the search algorithm, the search space and the child program. This decoupling makes it easy to change the search space and search algorithm (without and with weight sharing), as well as to add search capabilities to existing code and implement complex search flows. We then introduce PyGlove, a new Python library that implements this paradigm. Through case studies on ImageNet and NAS-Bench-101, we show that with PyGlove users can easily convert a static program into a search space, quickly iterate on the search spaces and search algorithms, and craft complex search flows to achieve better results. Copyright © 2021, The Authors. All rights reserved.

关键词： learning algorithms

Ethics lines and machine learning: A design and simulation of an association rules algorithm for exploiting the data

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Calvo, Patrici Egea-Moreno, Rebeca Universitat Jaume I

Data mining techniques offer great opportunities for developing ethics lines - tools for communication, participation and innovation whose main aim is to ensure improvements and compliance with the values, conduct and commitments making up the code of ethics. The aim of this study is to suggest a process for exploiting the data generated by the data generated and collected from an ethics line by extracting rules of association and applying the Apriori algorithm. This makes it possible to identify anomalies and behaviour patterns requiring action to review, correct, promote or expand them, as appropriate. Finally, I offer a simulated application of the Apriori algorithm, supplying it with synthetic data to find out its potential, strengths and limitations. © 2021, CC BY.

关键词： learning algorithms

Deep hedging of derivatives using reinforcement learning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Cao, Jay Chen, Jacky Hull, John Poulos, Zissis Joseph L. Rotman School of Management University of Toronto Canada

This paper shows how reinforcement learning can be used to derive optimal hedging strategies for derivatives when there are transaction costs. The paper illustrates the approach by showing the difference between using delta hedging and optimal hedging for a short position in a call option when the objective is to minimize a function equal to the mean hedging cost plus a constant times the standard deviation of the hedging cost. Two situations are considered. In the first, the asset price follows a geometric Brownian motion. In the second, the asset price follows a stochastic volatility process. The paper extends the basic reinforcement learning approach in a number of ways. First, it uses two different Q-functions so that both the expected value of the cost and the expected value of the square of the cost are tracked for different state/action combinations. This approach increases the range of objective functions that can be used. Second, it uses a learning algorithm that allows for continuous state and action space. Third, it compares the accounting P&L approach (where the hedged position is valued at each step) and the cash flow approach (where cash inflows and outflows are used). We find that a hybrid approach involving the use of an accounting P&L approach that incorporates a relatively simple valuation model works well. The valuation model does not have to correspond to the process assumed for the underlying asset price. © 2021, CC BY-NC-ND.

关键词： learning algorithms

Approximate Regions of Attraction in learning with Decision-Dependent Distributions

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Dong, Roy Zhang, Heling Ratliff, Lillian J. UIUC United States UW Canada

As data-driven methods are deployed in real-world settings, the processes that generate the observed data will often react to the decisions of the learner. For example, a data source may have some incentive for the algorithm to provide a particular label (e.g. approve a bank loan), and manipulate their features accordingly. Work in strategic classification and decision-dependent distributions seeks to characterize the closed-loop behavior of deploying learning algorithms by explicitly considering the effect of the classifier on the underlying data distribution. More recently, works in performative prediction seek to classify the closed-loop behavior by considering general properties of the mapping from classifier to data distribution, rather than an explicit form. Building on this notion, we analyze repeated risk minimization as the perturbed trajectories of the gradient flows of performative risk minimization. We consider the case where there may be multiple local minimizers of performative risk, motivated by situations where the initial conditions may have significant impact on the long-term behavior of the system. We provide sufficient conditions to characterize the region of attraction for the various equilibria in this settings. Additionally, we introduce the notion of performative alignment, which provides a geometric condition on the convergence of repeated risk minimization to performative risk minimizers. Copyright © 2021, The Authors. All rights reserved.

关键词： learning algorithms

Wavelet neural operator: a neural operator for parametric partial differential equations

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Tripura, Tapas Chakraborty, Souvik Department of Applied Mechanics Indian Institute of Technology Delhi India Indian Institute of Technology Delhi India

With massive advancements in sensor technologies and Internet-of-things (IoT), we now have access to terabytes of historical data;however, there is a lack of clarity in how to best exploit the data to predict future events. One possible alternative in this context is to utilize operator learning algorithm that directly learn nonlinear mapping between two functional spaces;this facilitates real-time prediction of naturally arising complex evolutionary dynamics. In this work, we introduce a novel operator learning algorithm referred to as the Wavelet Neural Operator (WNO) that blends integral kernel with wavelet transformation. WNO harnesses the superiority of the wavelets in time-frequency localization of the functions and enables accurate tracking of patterns in spatial domain and effective learning of the functional mappings. Since the wavelets are localized in both time/space and frequency, WNO can provide high spatial and frequency resolution. This offers learning of the finer details of the parametric dependencies in the solution for complex problems. The efficacy and robustness of the proposed WNO is illustrated on a wide array of problems involving Burger’s equation, Darcy flow, Navier-Stokes equation, Allen-Cahn equation, and Wave advection equation. Comparative study with respect to existing operator learning frameworks are presented. Finally, the proposed approach is used to build a digital twin capable of predicting Earth’s air temperature based on available historical data. © 2022, CC BY-NC-ND.

关键词： learning algorithms