检索结果-内蒙古大学图书馆

10th IEEE International Conference on Data Science and Advanced Analytics, DSAA 2023

作者： Ahmed, Nourhan Schmidt-Thieme, Lars University of Hildesheim Information Systems and Machine Learning Lab Hildesheim Germany

ISBN: (纸本)9798350345032

Remarkable progress has been achieved in generative modeling for time-series data with the introduction of Generative Adversarial Networks (GANs) [1]. GANs are neural networks that are meant to generate synthetic instances of data utilizing two neural networks, a generator and a discriminator, that operate against each other at the same time [1]. The generator learns to generate fake data to get the discriminator to classify its generated samples as authentic. The discriminator, on the other hand, attempts to distinguish between authentic and produced data. Finally, the generator could generate realistic data. GANs have demonstrated their ability to generate realistic data and have made remarkable progress in various tasks, such as the generation of time-series [4], images [5], and videos [3]. Particularly, a significant amount of work has utilized GANs based on Recurrent Neural Networks (RNNs) for time-series generation [4]. However, by carefully examining the generated samples from these models, we can observe that RNN-based GANs, such as LSTM GANs and gated recurrent GANs, cannot handle long sequences. Although RNN-based GANs can generate many realistic samples, there is still a difficulty in training due to exploding vanishing gradients and mode collapse that limits their generation capability. In addition, these RNN-based GANs are typically designed for regular time-series data, and thus cannot maintain informative varying intervals properly, which is a major concern for generating time-series *** this paper, we propose SparseGAN, a novel sparse self-attention-based GANs that allows for attention-driven, long-memory modeling for regular and irregular time-series generation through learned embedding space. This way, it can yield a more informative representation and capture long-range dependencies for time-series generation while using original data for supervision. SparseGAN comprises two essential sub-networks: the Supervision Network and the Generation Ne

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

GQFormer: A Multi-Quantile Generative Transformer for Time Series Forecasting

GQFormer: A Multi-Quantile Generative Transformer for Time S...

引用

2022 IEEE International Conference on Big Data, Big Data 2022

作者： Jawed, Shayan Schmidt-Thieme, Lars University of Hildesheim Information Systems and Machine Learning Lab Hildesheim Germany

ISBN: (纸本)9781665480451

We propose GQFormer, a probabilistic time series forecasting method that models the quantile function of the forecast distribution. Our methodology is rooted in the Implicit Quantile modeling approach, where samples from the Uniform distribution U( 0,1) are reparameterized to quantile values of the target distribution. This allows implicit generative quantile modeling without any prior assumptions on the data distribution like Gaussianity, common in prior works. Our work is distinguished from prior quantile forecasting methods by novel methodological advances that relate to directly modeling the correlations among multiple quantile estimations at each forecasting horizon. To this end, we firstly develop a parameters haring architecture that implicitly models multiple quantile estimations efficiently and secondly regularize these through a novel multi-task loss function formulation that optimizes for quantile estimations to be sharper estimations individually and on the whole be spread maximally apart to capture the various modes of the underlying distribution. We experimentally validate the superiority of the method to state-of-the-art probabilistic forecasting baselines and ablations to the loss formulation. © 2022 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

ProbSAINT: Probabilistic Tabular Regression for Used Car Pricing

ProbSAINT: Probabilistic Tabular Regression for Used Car Pri...

引用

2024 IEEE International Conference on Big Data, BigData 2024

作者： Madhusudhanan, Kiran Behrens, Gunnar Stubbemann, Maximilian Schmidt-Thieme, Lars University Of Hildesheim Information Systems and Machine Learning Lab Vwfs Data Analytics Research Center Germany Volkswagen Financial Services Ag Data Analytics & Ai Engineering United States

ISBN: (纸本)9798350362480

Used car pricing is a critical aspect of the automotive industry, influenced by many economic factors and market dynamics. With the recent surge in online marketplaces and increased demand for used cars, accurate pricing would benefit both buyers and sellers by ensuring fair transactions. However, the transition towards automated pricing algorithms using machine learning necessitates the comprehension of model uncertainties, specifically the ability to flag predictions that the model is unsure about. Although recent literature proposes the use of boosting algorithms or nearest neighbor-based approaches for swift and precise price predictions, encapsulating model uncertainties with such algorithms presents a complex challenge. We introduce ProbSAINT, a model that offers a principled approach for uncertainty quantification of its price predictions, along with accurate point predictions that are comparable to state-of-the-art boosting techniques. Furthermore, acknowledging that the business prefers pricing used cars based on the number of days the vehicle was listed for sale, we show how ProbSAINT can be used as a dynamic forecasting model for predicting price probabilities for different expected offer durations. Our experiments further indicate that ProbSAINT is especially accurate in instances where it is highly certain. This proves the applicability of its probabilistic predictions in real-world scenarios where trustworthiness is crucial. © 2024 IEEE.

关键词： Automotive industry

来源：评论

学校读者我要写书评

暂无评论

Deep Multi-Representation Model for Click-Through Rate Prediction

Deep Multi-Representation Model for Click-Through Rate Predi...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Shereen Elsayed Lars Schmidt-Thieme Information Systems and Machine Learning Lab University of Hildesheim Germany

Click-Through Rate prediction (CTR) is a crucial task for online advertising and recommender systems. Therefore, it has gained considerable attention in the past few years as it highly affects the revenue of several commercial platforms and online systems. The primary purpose of recent research emphasizes obtaining meaningful and powerful representations through mining low and high-feature interactions using various components such as Deep Neural Networks (DNN), CrossNets, or transformer blocks. However, models utilizing one representation for the input fields in each instance restrict the model's predictive power. Other models tend to be overly complicated to reach high input data expressiveness and predictive power. In this work, we propose a simple yet effective Deep Multi-Representation model (DeepMR) that is capable of learning informative representations by jointly training a mixture of two powerful feature representation learning components, namely DNNs and multi-head self-attentions. Furthermore, DeepMR integrates the novel residual with zero initialization (ReZero) connections to the DNN and the multi-head self-attention components for learning superior input representations. Experiments on three real-world datasets show that the proposed model significantly outperforms all state-of-the-art models with a relative improvement of up to 16.6% in the task of click-through rate prediction. Our implementation code and datasets are available here https://***/Shereen-Elsayed/DeepMR.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Sparse Self-Attention Guided Generative Adversarial Networks for Time-Series Generation

Sparse Self-Attention Guided Generative Adversarial Networks...

引用

International Conference on Data Science and Advanced Analytics (DSAA)

作者： Nourhan Ahmed Lars Schmidt-Thieme Information Systems and Machine Learning Lab University of Hildesheim Hildesheim Germany

关键词：

来源：评论

学校读者我要写书评

暂无评论

Probabilistic Forecasting of Irregularly Sampled Time Series with Missing Values via Conditional Normalizing Flows

arXiv

引用

arXiv 2024年

作者： Yalavarthi, Vijaya Krishna Scholz, Randolf Born, Stefan Schmidt-Thieme, Lars Information Systems and Machine Learning Lab University of Hildesheim Germany Institute of Mathematics TU Berlin Germany

Probabilistic forecasting of irregularly sampled multivariate time series with missing values is crucial for decision making in various domains, including health care, astronomy, and climate. State-of-the-art methods estimate only marginal distributions of observations in single channels and at single timepoints, assuming a Gaussian distribution for the data. In this work, we propose a novel model, ProFITi using conditional normalizing flows to learn multivariate conditional distribution: joint distribution of the future values of the time series conditioned on past observations and specific channels and timepoints, without assuming any fixed shape of the underlying distribution. As model components, we introduce a novel invertible triangular attention layer and an invertible non-linear activation function on and onto the whole real line. Through extensive experiments on 4 real-world datasets, ProFITi demonstrates significant improvement, achieving an average log-likelihood gain of 2.0 compared to the previous state-of-the-art method. © 2024, CC BY.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Robust Hyperbolic learning with Curvature-Aware Optimization

arXiv

引用

arXiv 2024年

作者： Bdeir, Ahmad Burchert, Johannes Schmidt-Thieme, Lars Landwehr, Niels Department of Data Analytics University of Hildesheim Germany The Information Systems and Machine Learning Lab University of Hildesheim Germany

Hyperbolic deep learning has become a growing research direction in computer vision due to the unique properties afforded by the alternate embedding space. The negative curvature and exponentially growing distance metric provide a natural framework for capturing hierarchical relationships between datapoints and allowing for finer separability between their embeddings. However, current hyperbolic learning approaches are still prone to overfitting, computationally expensive, and prone to instability, especially when attempting to learn the manifold curvature to adapt to tasks and different datasets. To address these issues, our paper presents a derivation for Riemannian AdamW that helps increase hyperbolic generalization ability. For improved stability, we introduce a novel fine-tunable hyperbolic scaling approach to constrain hyperbolic embeddings and reduce approximation errors. Using this along with our curvature-aware learning schema for Lorentzian Optimizers enables the combination of curvature and non-trivialized hyperbolic parameter learning. Our approach demonstrates consistent performance improvements across Computer Vision, EEG classification, and hierarchical metric learning tasks achieving state-of-the-art results in two domains and drastically reducing runtime. © 2024, CC BY.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Hyperparameter Tuning MLP's for Probabilistic Time Series Forecasting

arXiv

引用

arXiv 2024年

作者： Madhusudhanan, Kiran Jawed, Shayan Schmidt-Thieme, Lars Information Systems and Machine Learning Lab VWFS Data Analytics Research Center University of Hildesheim Hildesheim Germany

Time series forecasting attempts to predict future events by analyzing past trends and patterns. Although well researched, certain critical aspects pertaining to the use of deep learning in time series forecasting remain ambiguous. Our research primarily focuses on examining the impact of specific hyperparameters related to time series, such as context length and validation strategy, on the performance of the state-of-the-art MLP model in time series forecasting. We have conducted a comprehensive series of experiments involving 4800 configurations per dataset across 20 time series forecasting datasets, and our findings demonstrate the importance of tuning these parameters. Furthermore, in this work, we introduce the largest metadataset for time series forecasting to date, named TSBench, comprising 97200 evaluations, which is a twentyfold increase compared to previous works in the field. Finally, we demonstrate the utility of the created metadataset on multi-fidelity hyperparameter optimization tasks. © 2024, CC BY.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Context-Aware Sequential Model for Multi-Behaviour Recommendation

arXiv

引用

arXiv 2023年

作者： Elsayed, Shereen Rashed, Ahmed Schmidt-Thieme, Lars Information Systems and Machine Learning Lab University of Hildesheim Germany Volkswagen Financial Services Germany

Sequential recommendation models have recently become a crucial component for next-item recommendation tasks in various online platforms due to their unrivaled ability to capture complex sequential patterns in historical user interactions. Nevertheless, many recent sequential models mainly focus on modeling a single behavior, representing the platform’s target relation, e.g., purchase. While on the other hand, other implicit user interactions, such as click information, and add-to-favorite, can provide deeper insights into the users’ sequential behavior and allows better modeling of the users’ profiles. Recent work in multi-behavioral models has been trying to partially address this problem by focusing on utilizing graph-based approaches for modeling multi-behavior data as heterogeneous graphs. However, many fail or neglect to capture the sequential patterns simultaneously. While few recent time-aware multi-behavioral methods try to address both aspects at the same time, they still consider auxiliary behaviors of the same importance to the learning process, which might not be the case in many scenarios. In this work, we propose a Context-Aware Sequential Model (CASM) for multi-behavioral recommendations that leverages the advantages of sequential models and can support an arbitrary number of behaviors seamlessly. Specifically, context-aware multi-head self-attention layers are employed to capture the multi-behavior dependencies between the heterogeneous historical interactions. Furthermore, we utilize a weighted binary cross-entropy loss to weigh the different behaviors differently through the learning process of the model to allow more precise control of their contributions based on the target recommendation scenario. Experimental results on four real-world datasets show that the proposed model significantly outperforms multiple multi-behavioral and sequential recommendation state-of-the-art approaches. Copyright © 2023, The Authors. All rights reserved.

关键词： User profile

来源：评论

学校读者我要写书评

暂无评论

Deep Multi-Representation Model for Click-Through Rate Prediction

arXiv

引用

arXiv 2022年

作者： Elsayed, Shereen Schmidt-Thieme, Lars Information Systems and Machine Learning Lab Hildesheim Germany

Click-Through Rate prediction (CTR) is a crucial task in recommender systems, and it gained considerable attention in the past few years. The primary purpose of recent research emphasizes obtaining meaningful and powerful representations through mining low and high feature interactions using various components such as Deep Neural Networks (DNN), CrossNets, or transformer blocks. In this work, we propose the Deep Multi-Representation model (DeepMR) that jointly trains a mixture of two powerful feature representation learning components, namely DNNs and multi-head self-attentions. Furthermore, DeepMR integrates the novel residual with zero initialization (ReZero) connections to the DNN and the multi-head self-attention components for learning superior input representations. Experiments on three real-world datasets show that the proposed model significantly outperforms all state-of-the-art models in the task of click-through rate prediction. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：