arXiv subjects

Yinan Zhang

Publications and source records attributed to Yinan Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2023

Capturing Popularity Trends: A Simplistic Non-Personalized Approach for Enhanced Item Recommendation

Recommender systems have been gaining increasing research attention over the years. Most existing recommendation methods focus on capturing users' personalized preferences through historical user-item interactions, which may potentially violate user privacy. Additionally, these approaches often overlook the significance of the temporal fluctuation in item popularity that can sway users' decision-making. To bridge this gap, we propose Popularity-Aware Recommender (PARE), which makes non-personalized recommendations by predicting the items that will attain the highest popularity. PARE consists of four modules, each focusing on a different aspect: popularity history, temporal impact, periodic impact, and side information. Finally, an attention layer is leveraged to fuse the outputs of four modules. To our knowledge, this is the first work to explicitly model item popularity in recommendation systems. Extensive experiments show that PARE performs on par or even better than sophisticated state-of-the-art recommendation methods. Since PARE prioritizes item popularity over personalized user preferences, it can enhance existing recommendation methods as a complementary component. Our experiments demonstrate that integrating PARE with existing recommendation methods significantly surpasses the performance of standalone models, highlighting PARE's potential as a complement to existing recommendation methods. Furthermore, the simplicity of PARE makes it immensely practical for industrial applications and a valuable baseline for future research.

cs.IR↗

arXiv2023

Highly emissive, selective and omnidirectional thermal emitters mediated by machine learning for ultrahigh performance passive radiative cooling

Real-world passive radiative cooling requires highly emissive, selective, and omnidirectional thermal emitters to maintain the radiative cooler at a certain temperature below the ambient temperature while maximizing the net cooling power. Despite various selective thermal emitters have been demonstrated, it is still challenging to achieve these conditions simultaneously because of the extreme complexity of controlling thermal emission of photonic structures in multidimension. Here we demonstrated machine learning mediated hybrid metasurface thermal emitters with a high emissivity of ~0.92 within the atmospheric transparency window 8-13 μm, a large spectral selectivity of ~1.8 and a wide emission angle up to 80 degrees, simultaneously. This selective and omnidirectional thermal emitter has led to a new record of temperature reduction as large as ~15.4 degree under strong solar irradiation of ~800 W/m2, significantly surpassing the state-of-the-art results. The designed structures also show great potential in tackling the urban heat island effect, with modelling results suggesting a large energy saving and deployment area reduction. This research will make significant impact on passive radiative cooling, thermal energy photonics and tackling global climate change.

physics.app-ph↗

arXiv2022

Minimalist and High-performance Conversational Recommendation with Uncertainty Estimation for User Preference

Conversational recommendation system (CRS) is emerging as a user-friendly way to capture users' dynamic preferences over candidate items and attributes. Multi-shot CRS is designed to make recommendations multiple times until the user either accepts the recommendation or leaves at the end of their patience. Existing works are trained with reinforcement learning (RL), which may suffer from unstable learning and prohibitively high demands for computing. In this work, we propose a simple and efficient CRS, MInimalist Non-reinforced Interactive COnversational Recommender Network (MINICORN). MINICORN models the epistemic uncertainty of the estimated user preference and queries the user for the attribute with the highest uncertainty. The system employs a simple network architecture and makes the query-vs-recommendation decision using a single rule. Somewhat surprisingly, this minimalist approach outperforms state-of-the-art RL methods on three real-world datasets by large margins. We hope that MINICORN will serve as a valuable baseline for future research.

cs.IR↗

arXiv2022

Learning Large-scale Subsurface Simulations with a Hybrid Graph Network Simulator

Subsurface simulations use computational models to predict the flow of fluids (e.g., oil, water, gas) through porous media. These simulations are pivotal in industrial applications such as petroleum production, where fast and accurate models are needed for high-stake decision making, for example, for well placement optimization and field development planning. Classical finite difference numerical simulators require massive computational resources to model large-scale real-world reservoirs. Alternatively, streamline simulators and data-driven surrogate models are computationally more efficient by relying on approximate physics models, however they are insufficient to model complex reservoir dynamics at scale. Here we introduce Hybrid Graph Network Simulator (HGNS), which is a data-driven surrogate model for learning reservoir simulations of 3D subsurface fluid flows. To model complex reservoir dynamics at both local and global scale, HGNS consists of a subsurface graph neural network (SGNN) to model the evolution of fluid flows, and a 3D-U-Net to model the evolution of pressure. HGNS is able to scale to grids with millions of cells per time step, two orders of magnitude higher than previous surrogate models, and can accurately predict the fluid flow for tens of time steps (years into the future). Using an industry-standard subsurface flow dataset (SPE-10) with 1.1 million cells, we demonstrate that HGNS is able to reduce the inference time up to 18 times compared to standard subsurface simulators, and that it outperforms other learning-based models by reducing long-term prediction errors by up to 21%.

cs.LG↗

arXiv2021

Neural Radiance Flow for 4D View Synthesis and Video Processing

We present a method, Neural Radiance Flow (NeRFlow),to learn a 4D spatial-temporal representation of a dynamic scene from a set of RGB images. Key to our approach is the use of a neural implicit representation that learns to capture the 3D occupancy, radiance, and dynamics of the scene. By enforcing consistency across different modalities, our representation enables multi-view rendering in diverse dynamic scenes, including water pouring, robotic interaction, and real images, outperforming state-of-the-art methods for spatial-temporal view synthesis. Our approach works even when inputs images are captured with only one camera. We further demonstrate that the learned representation can serve as an implicit scene prior, enabling video processing tasks such as image super-resolution and de-noising without any additional supervision.

cs.CV↗

arXiv2021

Initialization Matters: Regularizing Manifold-informed Initialization for Neural Recommendation Systems

Proper initialization is crucial to the optimization and the generalization of neural networks. However, most existing neural recommendation systems initialize the user and item embeddings randomly. In this work, we propose a new initialization scheme for user and item embeddings called Laplacian Eigenmaps with Popularity-based Regularization for Isolated Data (LEPORID). LEPORID endows the embeddings with information regarding multi-scale neighborhood structures on the data manifold and performs adaptive regularization to compensate for high embedding variance on the tail of the data distribution. Exploiting matrix sparsity, LEPORID embeddings can be computed efficiently. We evaluate LEPORID in a wide range of neural recommendation models. In contrast to the recent surprising finding that the simple K-nearest-neighbor (KNN) method often outperforms neural recommendation systems, we show that existing neural systems initialized with LEPORID often perform on par or better than KNN. To maximize the effects of the initialization, we propose the Dual-Loss Residual Recommendation (DLR2) network, which, when initialized with LEPORID, substantially outperforms both traditional and state-of-the-art neural recommender systems.

cs.IR↗

arXiv2021

Image GANs meet Differentiable Rendering for Inverse Graphics and Interpretable 3D Neural Rendering

Differentiable rendering has paved the way to training neural networks to perform "inverse graphics" tasks such as predicting 3D geometry from monocular photographs. To train high performing models, most of the current approaches rely on multi-view imagery which are not readily available in practice. Recent Generative Adversarial Networks (GANs) that synthesize images, in contrast, seem to acquire 3D knowledge implicitly during training: object viewpoints can be manipulated by simply manipulating the latent codes. However, these latent codes often lack further physical interpretation and thus GANs cannot easily be inverted to perform explicit 3D reasoning. In this paper, we aim to extract and disentangle 3D knowledge learned by generative models by utilizing differentiable renderers. Key to our approach is to exploit GANs as a multi-view data generator to train an inverse graphics network using an off-the-shelf differentiable renderer, and the trained inverse graphics network as a teacher to disentangle the GAN's latent code into interpretable 3D properties. The entire architecture is trained iteratively using cycle consistency losses. We show that our approach significantly outperforms state-of-the-art inverse graphics networks trained on existing datasets, both quantitatively and via user studies. We further showcase the disentangled GAN as a controllable 3D "neural renderer", complementing traditional graphics renderers.

cs.CV↗

arXiv2021

Learning Hierarchical Review Graph Representations for Recommendation

The user review data have been demonstrated to be effective in solving different recommendation problems. Previous review-based recommendation methods usually employ sophisticated compositional models, such as Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN), to learn semantic representations from the review data for recommendation. However, these methods mainly capture the local dependency between neighbouring words in a word window, and they treat each review equally. Therefore, they may not be effective in capturing the global dependency between words, and tend to be easily biased by noise review information. In this paper, we propose a novel review-based recommendation model, named Review Graph Neural Network (RGNN). Specifically, RGNN builds a specific review graph for each individual user/item, which provides a global view about the user/item properties to help weaken the biases caused by noise review information. A type-aware graph attention mechanism is developed to learn semantic embeddings of words. Moreover, a personalized graph pooling operator is proposed to learn hierarchical representations of the review graph to form the semantic representation for each user/item. We compared RGNN with state-of-the-art review-based recommendation approaches on two real-world datasets. The experimental results indicate that RGNN consistently outperforms baseline methods, in terms of Mean Square Error (MSE).

cs.IR↗

arXiv2020

PuzzleFlex: kinematic motion of chains with loose joints

This paper presents a method of computing free motions of a planar assembly of rigid bodies connected by loose joints. Joints are modeled using local distance constraints, which are then linearized with respect to configuration space velocities, yielding a linear programming formulation that allows analysis of systems with thousands of rigid bodies. Potential applications include analysis of collections of modular robots, structural stability perturbation analysis, tolerance analysis for mechanical systems, and formation control of mobile robots.

cs.RO↗

arXiv2020

Reductive Clustering: An Efficient Linear-time Graph-based Divisive Cluster Analysis Approach

We propose an efficient linear-time graph-based divisive cluster analysis approach called Reductive Clustering. The approach tries to reveal the hierarchical structural information through reducing the graph into a more concise one repeatedly. With the reductions, the original graph can be divided into subgraphs recursively, and a lite informative dendrogram is constructed based on the divisions. The reduction consists of three steps: selection, connection, and partition. First a subset of vertices of the graph are selected as representatives to build a concise graph. The representatives are re-connected to maintain a consistent structure with the previous graph. If possible, the concise graph is divided into subgraphs, and each subgraph is further reduced recursively until the termination condition is met. We discuss the approach, along with several selection and connection methods, in detail both theoretically and experimentally in this paper. Our implementations run in linear time and achieve outstanding performance on various types of datasets. Experimental results show that they outperform state-of-the-art clustering algorithms with significantly less computing resource requirements.

cs.AI↗

arXiv2020

Topological-darkness-assisted phase regulation for atomically thin meta-optics

Two-dimensional (2D) noble-metal dichalcogenides have emerged as a new platform for the realization of versatile flat optics with a considerable degree of miniaturization. However, light field manipulation at the atomic scale is widely considered unattainable since the vanishing thickness and intrinsic losses of 2D materials completely suppress both resonances and phase accumulation effects. Empowered by conventionally perceived adverse effects of intrinsic losses, we show that the structured PtSe2 films integrated with a uniform substrate can regulate nontrivial singular phase and realize atomic-thick meta-optics in the presence of topological darkness. We experimentally demonstrate a series of atomic-thick binary meta-optics that allows angle-robust and high unit-thickness diffraction efficiency of 0.96%/nm in visible frequencies, given its thickness of merely 4.3 nm. Our results unlock the potential of a new class of 2D flat optics for light field manipulation at an atomic thickness.

physics.optics↗

arXiv2020

Towards a Soft Faceted Browsing Scheme for Information Access

Faceted browsing is a commonly supported feature of user interfaces for access to information. Existing interfaces generally treat facet values selected by a user as hard filters and respond to the user by only displaying information items strictly satisfying the filters and in their original ranking order. We propose a novel alternative strategy for faceted browsing, called soft faceted browsing, where the system also includes some possibly relevant items outside the selected filter in a non-intrusive way and re-ranks the items to better satisfy the user's information need. Such a soft faceted browsing strategy can be beneficial when the user does not have a very confident and strict preference for the selected facet values, and is especially appropriate for applications such as e-commerce search where the user would like to explore a larger space before finalizing a purchasing decision. We propose a probabilistic framework for modeling and solving the soft faceted browsing problem, and apply the framework to study the case of facet filter selection in e-commerce search engines. Preliminary experiment results demonstrate the soft faceted browsing scheme is better than the traditional faceted browsing scheme in terms of its efficiency in helping users navigate in the information space.

cs.IR↗

arXiv2019

Explicit Pairwise Word Interaction Modeling Improves Pretrained Transformers for English Semantic Similarity Tasks

In English semantic similarity tasks, classic word embedding-based approaches explicitly model pairwise "interactions" between the word representations of a sentence pair. Transformer-based pretrained language models disregard this notion, instead modeling pairwise word interactions globally and implicitly through their self-attention mechanism. In this paper, we hypothesize that introducing an explicit, constrained pairwise word interaction mechanism to pretrained language models improves their effectiveness on semantic similarity tasks. We validate our hypothesis using BERT on four tasks in semantic textual similarity and answer sentence selection. We demonstrate consistent improvements in quality by adding an explicit pairwise word interaction module to BERT.

cs.CL↗

arXiv2019

Diversity-Promoting Deep Reinforcement Learning for Interactive Recommendation

Interactive recommendation that models the explicit interactions between users and the recommender system has attracted a lot of research attentions in recent years. Most previous interactive recommendation systems only focus on optimizing recommendation accuracy while overlooking other important aspects of recommendation quality, such as the diversity of recommendation results. In this paper, we propose a novel recommendation model, named \underline{D}iversity-promoting \underline{D}eep \underline{R}einforcement \underline{L}earning (D$^2$RL), which encourages the diversity of recommendation results in interaction recommendations. More specifically, we adopt a Determinantal Point Process (DPP) model to generate diverse, while relevant item recommendations. A personalized DPP kernel matrix is maintained for each user, which is constructed from two parts: a fixed similarity matrix capturing item-item similarity, and the relevance of items dynamically learnt through an actor-critic reinforcement learning framework. We performed extensive offline experiments as well as simulated online experiments with real world datasets to demonstrate the effectiveness of the proposed model.

cs.IR↗

arXiv2019

Towards Physically Safe Reinforcement Learning under Supervision

This paper addresses the question of how a previously available control policy $π_s$ can be used as a supervisor to more quickly and safely train a new learned control policy $π_L$ for a robot. A weighted average of the supervisor and learned policies is used during trials, with a heavier weight initially on the supervisor, in order to allow safe and useful physical trials while the learned policy is still ineffective. During the process, the weight is adjusted to favor the learned policy. As weights are adjusted, the learned network must compensate so as to give safe and reasonable outputs under the different weights. A pioneer network is introduced that pre-learns a policy that performs similarly to the current learned policy under the planned next step for new weights; this pioneer network then replaces the currently learned network in the next set of trials. Experiments in OpenAI Gym demonstrate the effectiveness of the proposed method.

cs.LG↗

arXiv2017

Valuations of $p$-adic regulators of cyclic cubic fields

We compute the $p$-adic regulator of cyclic cubic extensions of $\mathbb Q$ with discriminant up to $10^{16}$ for $3<p<100$, and observe the distribution of the $p$-adic valuation of the regulators. We find that for almost all primes, the observation matches the model that the entries in the regulator matrix are random elements with respect to the obvious restrictions. Based on this random matrix model, a conjecture on the distribution of the valuations of $p$-adic regulators of cyclic cubic fields is stated.

math.NT↗

arXiv2012

Broadband Reconfiguration of OptoMechanical Filters

We demonstrate broad-band reconfiguration of coupled photonic crystal nanobeam cavities by using optical gradient force induced mechanical actuation. Propagating waveguide modes that exist over wide wavelength range are used to actuate the structures and in that way control the resonance of localized cavity mode. Using this all-optical approach, more than 18 linewidths of tuning range is demonstrated. Using on-chip temperature self-referencing method that we developed, we determined that 20 % of the total tuning was due to optomechanical reconfiguration and the rest due to thermo-optic effects. Independent control of mechanical and optical resonances of our structures, by means of optical stiffening, is also demonstrated.

physics.optics↗

arXiv2011

High-Q/V air-mode photonic crystal cavities at microwave frequencies

We present results for a photonic microwave resonator designed and fabricated at 17.4GHz with a record high Quality factor (Q = 26,400) at room temperature over a mode volume smaller than one cubic wavelength. The cavity is uniquely designed to have its electric field concentrated in air, which allows for efficient coupling to free space and facilitates interactions with gaseous atomic systems and fluids.

physics.optics↗