January 30, 2020

2753 words 13 mins read

Paper Group ANR 446

Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language. Generalization in Transfer Learning. Model-based Lookahead Reinforcement Learning. Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit. KBSET – Knowledge-Based Support for Scholarly Editing and Text Processing. Three-Way Decisions-B …

Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language


Title	Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language
Authors	Yuri Kuratov, Mikhail Arkhipov
Abstract	The paper introduces methods of adaptation of multilingual masked language models for a specific language. Pre-trained bidirectional language models show state-of-the-art performance on a wide range of tasks including reading comprehension, natural language inference, and sentiment analysis. At the moment there are two alternative approaches to train such models: monolingual and multilingual. While language specific models show superior performance, multilingual models allow to perform a transfer from one language to another and solve tasks for different languages simultaneously. This work shows that transfer learning from a multilingual model to monolingual model results in significant growth of performance on such tasks as reading comprehension, paraphrase detection, and sentiment analysis. Furthermore, multilingual initialization of monolingual model substantially reduces training time. Pre-trained models for the Russian language are open sourced.
Tasks	Natural Language Inference, Reading Comprehension, Sentiment Analysis, Transfer Learning
Published	2019-05-17
URL	https://arxiv.org/abs/1905.07213v1
PDF	https://arxiv.org/pdf/1905.07213v1.pdf
PWC	https://paperswithcode.com/paper/adaptation-of-deep-bidirectional-multilingual
Repo
Framework

Generalization in Transfer Learning


Title	Generalization in Transfer Learning
Authors	Suzan Ece Ada, Emre Ugur, H. Levent Akin
Abstract	Agents trained with deep reinforcement learning algorithms are capable of performing highly complex tasks including locomotion in continuous environments. In order to attain a human-level performance, the next step of research should be to investigate the ability to transfer the learning acquired in one task to a different set of tasks. Concerns on generalization and overfitting in deep reinforcement learning are not usually addressed in current transfer learning research. This issue results in underperforming benchmarks and inaccurate algorithm comparisons due to rudimentary assessments. In this study, we primarily propose regularization techniques in deep reinforcement learning for continuous control through the application of sample elimination and early stopping. First, the importance of the inclusion of training iteration to the hyperparameters in deep transfer learning problems will be emphasized. Because source task performance is not indicative of the generalization capacity of the algorithm, we start by proposing various transfer learning evaluation methods that acknowledge the training iteration as a hyperparameter. In line with this, we introduce an additional step of resorting to earlier snapshots of policy parameters depending on the target task due to overfitting to the source task. Then, in order to generate robust policies,we discard the samples that lead to overfitting via strict clipping. Furthermore, we increase the generalization capacity in widely used transfer learning benchmarks by using entropy bonus, different critic methods and curriculum learning in an adversarial setup. Finally, we evaluate the robustness of these techniques and algorithms on simulated robots in target environments where the morphology of the robot, gravity and tangential friction of the environment are altered from the source environment.
Tasks	Continuous Control, Transfer Learning
Published	2019-09-03
URL	https://arxiv.org/abs/1909.01331v1
PDF	https://arxiv.org/pdf/1909.01331v1.pdf
PWC	https://paperswithcode.com/paper/generalization-in-transfer-learning
Repo
Framework

Model-based Lookahead Reinforcement Learning


Title	Model-based Lookahead Reinforcement Learning
Authors	Zhang-Wei Hong, Joni Pajarinen, Jan Peters
Abstract	Model-based Reinforcement Learning (MBRL) allows data-efficient learning which is required in real world applications such as robotics. However, despite the impressive data-efficiency, MBRL does not achieve the final performance of state-of-the-art Model-free Reinforcement Learning (MFRL) methods. We leverage the strengths of both realms and propose an approach that obtains high performance with a small amount of data. In particular, we combine MFRL and Model Predictive Control (MPC). While MFRL’s strength in exploration allows us to train a better forward dynamics model for MPC, MPC improves the performance of the MFRL policy by sampling-based planning. The experimental results in standard continuous control benchmarks show that our approach can achieve MFRL`s level of performance while being as data-efficient as MBRL. \|
Tasks	Continuous Control
Published	2019-08-15
URL	https://arxiv.org/abs/1908.06012v1
PDF	https://arxiv.org/pdf/1908.06012v1.pdf
PWC	https://paperswithcode.com/paper/model-based-lookahead-reinforcement-learning
Repo
Framework

Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit


Title	Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit
Authors	Song Mei, Theodor Misiakiewicz, Andrea Montanari
Abstract	We consider learning two layer neural networks using stochastic gradient descent. The mean-field description of this learning dynamics approximates the evolution of the network weights by an evolution in the space of probability distributions in $R^D$ (where $D$ is the number of parameters associated to each neuron). This evolution can be defined through a partial differential equation or, equivalently, as the gradient flow in the Wasserstein space of probability distributions. Earlier work shows that (under some regularity assumptions), the mean field description is accurate as soon as the number of hidden units is much larger than the dimension $D$. In this paper we establish stronger and more general approximation guarantees. First of all, we show that the number of hidden units only needs to be larger than a quantity dependent on the regularity properties of the data, and independent of the dimensions. Next, we generalize this analysis to the case of unbounded activation functions, which was not covered by earlier bounds. We extend our results to noisy stochastic gradient descent. Finally, we show that kernel ridge regression can be recovered as a special limit of the mean field analysis.
Tasks
Published	2019-02-16
URL	http://arxiv.org/abs/1902.06015v1
PDF	http://arxiv.org/pdf/1902.06015v1.pdf
PWC	https://paperswithcode.com/paper/mean-field-theory-of-two-layers-neural
Repo
Framework

KBSET – Knowledge-Based Support for Scholarly Editing and Text Processing


Title	KBSET – Knowledge-Based Support for Scholarly Editing and Text Processing
Authors	Jana Kittelmann, Christoph Wernhard
Abstract	KBSET supports a practical workflow for scholarly editing, based on using LaTeX with dedicated commands for semantics-oriented markup and a Prolog-implemented core system. Prolog plays there various roles: as query language and access mechanism for large Semantic Web fact bases, as data representation of structured documents and as a workflow model for advanced application tasks. The core system includes a LaTeX parser and a facility for the identification of named entities. We also sketch future perspectives of this approach to scholarly editing based on techniques of computational logic.
Tasks
Published	2019-08-29
URL	https://arxiv.org/abs/1908.11135v1
PDF	https://arxiv.org/pdf/1908.11135v1.pdf
PWC	https://paperswithcode.com/paper/kbset-knowledge-based-support-for-scholarly
Repo
Framework

Three-Way Decisions-Based Conflict Analysis Models


Title	Three-Way Decisions-Based Conflict Analysis Models
Authors	Guangming Lang
Abstract	Three-way decision theory, which trisects the universe with less risks or costs, is considered as a powerful mathematical tool for handling uncertainty in incomplete and imprecise information tables, and provides an effective tool for conflict analysis decision making in real-time situations. In this paper, we propose the concepts of the agreement, disagreement and neutral subsets of a strategy with two evaluation functions, which establish the three-way decisions-based conflict analysis models(TWDCAMs) for trisecting the universe of agents, and employ a pair of two-way decisions models to interpret the mechanism of the three-way decision rules for an agent. Subsequently, we develop the concepts of the agreement, disagreement and neutral strategies of an agent group with two evaluation functions, which build the TWDCAMs for trisecting the universe of issues, and take a couple of two-way decisions models to explain the mechanism of the three-way decision rules for an issue. Finally, we reconstruct Fan, Qi and Wei’s conflict analysis models(FQWCAMs) and Sun, Ma and Zhao’s conflict analysis models(SMZCAMs) with two evaluation functions, and interpret FQWCAMs and SMZCAMs with a pair of two-day decisions models, which illustrates that FQWCAMs and SMZCAMs are special cases of TWDCAMs.
Tasks	Decision Making
Published	2019-03-07
URL	http://arxiv.org/abs/1903.03205v1
PDF	http://arxiv.org/pdf/1903.03205v1.pdf
PWC	https://paperswithcode.com/paper/three-way-decisions-based-conflict-analysis
Repo
Framework

A Robust Two-Sample Test for Time Series data


Title	A Robust Two-Sample Test for Time Series data
Authors	Alexis Bellot, Mihaela van der Schaar
Abstract	We develop a general framework for hypothesis testing with time series data. The problem is to distinguish between the mean functions of the underlying temporal processes of populations of times series, which are often irregularly sampled and measured with error. Such an observation pattern can result in substantial uncertainty about the underlying trajectory, quantifying it accurately is important to ensure robust tests. We propose a new test statistic that views each trajectory as a sample from a distribution on functions and considers the distributions themselves to encode the uncertainty between observations. We derive asymptotic null distributions and power functions for our test and put emphasis on computational considerations by giving an efficient kernel learning framework to prevent over-fitting in small samples and also showing how to scale our test to densely sampled time series. We conclude with performance evaluations on synthetic data and experiments on healthcare and climate change data.
Tasks	Time Series
Published	2019-07-09
URL	https://arxiv.org/abs/1907.04081v1
PDF	https://arxiv.org/pdf/1907.04081v1.pdf
PWC	https://paperswithcode.com/paper/a-robust-two-sample-test-for-time-series-data
Repo
Framework

Deep Invertible Networks for EEG-based brain-signal decoding


Title	Deep Invertible Networks for EEG-based brain-signal decoding
Authors	Robin Tibor Schirrmeister, Tonio Ball
Abstract	In this manuscript, we investigate deep invertible networks for EEG-based brain signal decoding and find them to generate realistic EEG signals as well as classify novel signals above chance. Further ideas for their regularization towards better decoding accuracies are discussed.
Tasks	EEG
Published	2019-07-17
URL	https://arxiv.org/abs/1907.07746v1
PDF	https://arxiv.org/pdf/1907.07746v1.pdf
PWC	https://paperswithcode.com/paper/deep-invertible-networks-for-eeg-based-brain
Repo
Framework

The relationship between trust in AI and trustworthy machine learning technologies


Title	The relationship between trust in AI and trustworthy machine learning technologies
Authors	Ehsan Toreini, Mhairi Aitken, Kovila Coopamootoo, Karen Elliott, Carlos Gonzalez Zelaya, Aad van Moorsel
Abstract	To build AI-based systems that users and the public can justifiably trust one needs to understand how machine learning technologies impact trust put in these services. To guide technology developments, this paper provides a systematic approach to relate social science concepts of trust with the technologies used in AI-based services and products. We conceive trust as discussed in the ABI (Ability, Benevolence, Integrity) framework and use a recently proposed mapping of ABI on qualities of technologies. We consider four categories of machine learning technologies, namely these for Fairness, Explainability, Auditability and Safety (FEAS) and discuss if and how these possess the required qualities. Trust can be impacted throughout the life cycle of AI-based systems, and we introduce the concept of Chain of Trust to discuss technological needs for trust in different stages of the life cycle. FEAS has obvious relations with known frameworks and therefore we relate FEAS to a variety of international Principled AI policy and technology frameworks that have emerged in recent years.
Tasks
Published	2019-11-27
URL	https://arxiv.org/abs/1912.00782v2
PDF	https://arxiv.org/pdf/1912.00782v2.pdf
PWC	https://paperswithcode.com/paper/the-relationship-between-trust-in-ai-and
Repo
Framework

PointIT: A Fast Tracking Framework Based on 3D Instance Segmentation


Title	PointIT: A Fast Tracking Framework Based on 3D Instance Segmentation
Authors	Yuan Wang, Yang Yu, Ming Liu
Abstract	Recently most popular tracking frameworks focus on 2D image sequences. They seldom track the 3D object in point clouds. In this paper, we propose PointIT, a fast, simple tracking method based on 3D on-road instance segmentation. Firstly, we transform 3D LiDAR data into the spherical image with the size of 64 x 512 x 4 and feed it into instance segment model to get the predicted instance mask for each class. Then we use MobileNet as our primary encoder instead of the original ResNet to reduce the computational complexity. Finally, we extend the Sort algorithm with this instance framework to realize tracking in the 3D LiDAR point cloud data. The model is trained on the spherical images dataset with the corresponding instance label masks which are provided by KITTI 3D Object Track dataset. According to the experiment results, our network can achieve on Average Precision (AP) of 0.617 and the performance of multi-tracking task has also been improved.
Tasks	3D Instance Segmentation, Instance Segmentation, Semantic Segmentation
Published	2019-02-18
URL	http://arxiv.org/abs/1902.06379v1
PDF	http://arxiv.org/pdf/1902.06379v1.pdf
PWC	https://paperswithcode.com/paper/pointit-a-fast-tracking-framework-based-on-3d
Repo
Framework

What is needed for simple spatial language capabilities in VQA?


Title	What is needed for simple spatial language capabilities in VQA?
Authors	Alexander Kuhnle, Ann Copestake
Abstract	Visual question answering (VQA) comprises a variety of language capabilities. The diagnostic benchmark dataset CLEVR has fueled progress by helping to better assess and distinguish models in basic abilities like counting, comparing and spatial reasoning in vitro. Following this approach, we focus on spatial language capabilities and investigate the question: what are the key ingredients to handle simple visual-spatial relations? We look at the SAN, RelNet, FiLM and MC models and evaluate their learning behavior on diagnostic data which is solely focused on spatial relations. Via comparative analysis and targeted model modification we identify what really is required to substantially improve upon the CNN-LSTM baseline.
Tasks	Question Answering, Visual Question Answering
Published	2019-08-17
URL	https://arxiv.org/abs/1908.06336v2
PDF	https://arxiv.org/pdf/1908.06336v2.pdf
PWC	https://paperswithcode.com/paper/what-is-needed-for-simple-spatial-language
Repo
Framework

Domain Aggregation Networks for Multi-Source Domain Adaptation


Title	Domain Aggregation Networks for Multi-Source Domain Adaptation
Authors	Junfeng Wen, Russell Greiner, Dale Schuurmans
Abstract	In many real-world applications, we want to exploit multiple source datasets of similar tasks to learn a model for a different but related target dataset – e.g., recognizing characters of a new font using a set of different fonts. While most recent research has considered ad-hoc combination rules to address this problem, we extend previous work on domain discrepancy minimization to develop a finite-sample generalization bound, and accordingly propose a theoretically justified optimization procedure. The algorithm we develop, Domain AggRegation Network (DARN), is able to effectively adjust the weight of each source domain during training to ensure relevant domains are given more importance for adaptation. We evaluate the proposed method on real-world sentiment analysis and digit recognition datasets and show that DARN can significantly outperform the state-of-the-art alternatives.
Tasks	Domain Adaptation, Sentiment Analysis
Published	2019-09-11
URL	https://arxiv.org/abs/1909.05352v2
PDF	https://arxiv.org/pdf/1909.05352v2.pdf
PWC	https://paperswithcode.com/paper/domain-aggregation-networks-for-multi-source
Repo
Framework

Performance study of distributed Apriori-like frequent itemsets mining


Title	Performance study of distributed Apriori-like frequent itemsets mining
Authors	Lamine M. Aouad, Nhien-An Le-Khac, Tahar M. Kechadi
Abstract	In this article, we focus on distributed Apriori-based frequent itemsets mining. We present a new distributed approach which takes into account inherent characteristics of this algorithm. We study the distribution aspect of this algorithm and give a comparison of the proposed approach with a classical Apriori-like distributed algorithm, using both analytical and experimental studies. We find that under a wide range of conditions and datasets, the performance of a distributed Apriori-like algorithm is not related to global strategies of pruning since the performance of the local Apriori generation is usually characterized by relatively high success rates of candidate sets frequency at low levels which switch to very low rates at some stage, and often drops to zero. This means that the intermediate communication steps and remote support counts computation and collection in classical distributed schemes are computationally inefficient locally, and then constrains the global performance. Our performance evaluation is done on a large cluster of workstations using the Condor system and its workflow manager DAGMan. The results show that the presented approach greatly enhances the performance and achieves good scalability compared to a typical distributed Apriori founded algorithm.
Tasks
Published	2019-02-21
URL	http://arxiv.org/abs/1903.03008v1
PDF	http://arxiv.org/pdf/1903.03008v1.pdf
PWC	https://paperswithcode.com/paper/performance-study-of-distributed-apriori-like
Repo
Framework

ML for Flood Forecasting at Scale


Title	ML for Flood Forecasting at Scale
Authors	Sella Nevo, Vova Anisimov, Gal Elidan, Ran El-Yaniv, Pete Giencke, Yotam Gigi, Avinatan Hassidim, Zach Moshe, Mor Schlesinger, Guy Shalev, Ajai Tirumali, Ami Wiesel, Oleg Zlydenko, Yossi Matias
Abstract	Effective riverine flood forecasting at scale is hindered by a multitude of factors, most notably the need to rely on human calibration in current methodology, the limited amount of data for a specific location, and the computational difficulty of building continent/global level models that are sufficiently accurate. Machine learning (ML) is primed to be useful in this scenario: learned models often surpass human experts in complex high-dimensional scenarios, and the framework of transfer or multitask learning is an appealing solution for leveraging local signals to achieve improved global performance. We propose to build on these strengths and develop ML systems for timely and accurate riverine flood prediction.
Tasks	Calibration
Published	2019-01-28
URL	http://arxiv.org/abs/1901.09583v1
PDF	http://arxiv.org/pdf/1901.09583v1.pdf
PWC	https://paperswithcode.com/paper/ml-for-flood-forecasting-at-scale
Repo
Framework

ABCDP: Approximate Bayesian Computation Meets Differential Privacy


Title	ABCDP: Approximate Bayesian Computation Meets Differential Privacy
Authors	Mijung Park, Wittawat Jitkrittum
Abstract	We develop a novel approximate Bayesian computation (ABC) framework, ABCDP, that obeys the notion of differential privacy (DP). Under our framework, simply performing ABC inference with a mild modification yields differentially private posterior samples. We theoretically analyze the interplay between the ABC similarity threshold $\epsilon_{abc}$ (for comparing the similarity between real and simulated data) and the resulting privacy level $\epsilon_{dp}$ of the posterior samples, in two types of frequently-used ABC algorithms. We apply ABCDP to simulated data as well as privacy-sensitive real data. The results suggest that tuning the similarity threshold $\epsilon_{abc}$ helps us obtain better privacy and accuracy trade-off.
Tasks
Published	2019-10-11
URL	https://arxiv.org/abs/1910.05103v1
PDF	https://arxiv.org/pdf/1910.05103v1.pdf
PWC	https://paperswithcode.com/paper/abcdp-approximate-bayesian-computation-meets
Repo
Framework