Bo Li

Shaping in reinforcement learning via knowledge transferred from human-demonstrations

2015

Transfer has been widely used to ameliorate the slow convergence speed of reinforcement learning (RL) by reusing the previous obtained knowledge from other related but distinct tasks. In this paper, we propose a framework to …

Measuring Service Utilities in Service Value Networks

2017 · International Journal of Information Systems in the Service Sector

In spite of the importance of service value networks (SVNs) in today's service sectors, academic studies of SVNs, in terms of their formalisms, models and value creation processes, are still lacking, with only sporadic publications …

Robust Physical-World Attacks on Machine Learning Models.

2017 · arXiv (Cornell University)

Deep neural network-based classifiers are known to be vulnerable to adversarial examples that can fool them into misclassifying their input through the addition of small-magnitude perturbations. However, recent studies have demonstrated that such adversarial examples …

Generating Adversarial Examples with Adversarial Networks

2018 · arXiv (Cornell University)

Deep neural networks (DNNs) have been found to be vulnerable to adversarial examples resulting from adding small-magnitude perturbations to inputs. Such adversarial examples can mislead DNNs to produce adversary-selected results. Different attack strategies have been …

Adversarial Texts with Gradient Methods

2018 · arXiv (Cornell University)

Adversarial samples for images have been extensively studied in the literature. Among many of the attacking methods, gradient-based methods are both effective and easy to compute. In this work, we propose a framework to adapt …

Generative Model: Membership Attack,Generalization and Diversity

2018 · arXiv (Cornell University)

In this paper we propose a new membership attack method called co-membership attacks against deep generative models including Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs). Specifically, membership attack aims to check whether a given …

A Precise Tidal Level Prediction Method Using Improved Extreme Learning Machine with Sliding Data Window

2018

To improve the tidal prediction accuracy, a prediction scheme is proposed by using improved Extreme Learning Machine (IELM) based on a sliding data window. The changes of tidal level are complex processes which are influenced …

Learning Neural Representation for CLIR with Adversarial Framework

2018

The existing studies in cross-language information retrieval (CLIR) mostly rely on general text representation models (e.g., vector space model or latent semantic analysis). These models are not optimized for the target retrieval task. In this …

Robust Inference via Generative Classifiers for Handling Noisy Labels

2019 · arXiv (Cornell University)

Large-scale datasets may contain significant proportions of noisy (incorrect) class labels, and it is well-known that modern deep neural networks (DNNs) poorly generalize from such noisy training datasets. To mitigate the issue, we propose a …

Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality

2018 · Own your potential (DEAKIN)

© Learning Representations, ICLR 2018 - Conference Track Proceedings.All right reserved. Deep Neural Networks (DNNs) have recently been shown to be vulnerable against adversarial examples, which are carefully crafted instances that can mislead DNNs to …

Personalized exercise recommendation via implicit skills

2019 · Proceedings of the ACM Turing Celebration Conference - China

Cognitive diagnosis methods need to assess the students' skills to provide personalized exercise recommendation. To perform this assessment, an initially hand built Q-matrix are presented to students, which would affect the recommendation results in intelligence …

Improving Certified Robustness via Statistical Learning with Logical Reasoning

2020 · arXiv (Cornell University)

Intensive algorithmic efforts have been made to enable the rapid improvements of certificated robustness for complex ML models recently. However, current robustness certification methods are only able to certify under a limited perturbation radius. Given …

T3: Tree-Autoencoder Constrained Adversarial Text Generation for Targeted Attack

2019 · arXiv (Cornell University)

Adversarial attacks against natural language processing systems, which perform seemingly innocuous modifications to inputs, can induce arbitrary mistakes to the target models. Though raised great concerns, such adversarial attacks can be leveraged to estimate the …

Research on One-Dimensional Barcode Recognition of Express Delivery Based on Machine Vision

2020 · CICTP 2020

To improve the efficiency and accuracy of barcode recognition in express delivery on the production line, a one-dimensional barcode recognition algorithm based on machine vision technology was studied. This studies first express image acquisition platform …

InfoBERT: Improving Robustness of Language Models from An Information\n Theoretic Perspective

2020 · arXiv (Cornell University)

Large-scale language models such as BERT have achieved state-of-the-art\nperformance across a wide range of NLP tasks. Recent studies, however, show\nthat such BERT-based models are vulnerable facing the threats of textual\nadversarial attacks. We aim to address …

What Would Jiminy Cricket Do? Towards Agents That Behave Morally

2021 · arXiv (Cornell University)

When making everyday decisions, people are guided by their conscience, an internal sense of right and wrong. By contrast, artificial agents are currently not endowed with a moral sense. As a consequence, they may learn …

Global Convergence of MAML and Theory-Inspired Neural Architecture Search for Few-Shot Learning

2022 · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Model-agnostic meta-learning (MAML) and its variants have become popular approaches for few-shot learning. However, due to the non-convexity of deep neural nets (DNNs) and the bi-level formulation of MAML, the theoretical properties of MAML with …

Hybrid Anomaly Detection via Multihead Dynamic Graph Attention Networks for Multivariate Time Series

2022 · IEEE Access

In the real world, a large number of multivariate time series data are generated by Internet of Things systems, which are composed of many connected sensing devices. Therefore, it is impractical to consider only a …

SecretGen: Privacy Recovery on Pre-Trained Models via Distribution Discrimination

2022 · arXiv (Cornell University)

Transfer learning through the use of pre-trained models has become a growing trend for the machine learning community. Consequently, numerous pre-trained models are released online to facilitate further research. However, it raises extensive concerns on …

Elderly Nursing Service System Based on Improved Particle Swarm Optimization Algorithm for Camera Calibration

2022

It is the obligation of future generations to ensure that the elderly can enjoy their old age happily and comfortably. It is also a manifestation of social civilization and progress, and it is the inheritance …

Modular Domain Adaptation for Conformer-Based Streaming ASR

2023 · arXiv (Cornell University)

Speech data from different domains has distinct acoustic and linguistic characteristics. It is common to train a single multidomain model such as a Conformer transducer for speech recognition on a mixture of data from all …

Adaptive Graph Representation Learning for Next POI Recommendation

2023

Next Point-of-Interest (POI) recommendation is an essential part of the flourishing location-based applications, where the demands of users are not only conditioned by their recent check-in behaviors but also by the critical influence stemming from …

Multiferroic Magnon Spin-Torque Based Reconfigurable Logic-In-Memory

2023 · arXiv (Cornell University)

Magnons, bosonic quasiparticles carrying angular momentum, can flow through insulators for information transmission with minimal power dissipation. However, it remains challenging to develop a magnon-based logic due to the lack of efficient electrical manipulation of …

PV power prediction method based on ground-based cloud map and hybrid neural network

2023

Under cloudy weather, PV power will fluctuate dramatic due to the sporty cloud shading. However, most existing PV power prediction models do not utilize the cloud map sufficiently. In this paper, we propose a novel …

CBD: A Certified Backdoor Detector Based on Local Dominant Probability

2023 · arXiv (Cornell University)

Backdoor attack is a common threat to deep neural networks. During testing, samples embedded with a backdoor trigger will be misclassified as an adversarial target by a backdoored model, while samples without the backdoor trigger …

Efficient retrieval of power structured data with global data access view

2023

With the deepening of power grid informatization construction, the number of structured data such as equipment, network and operation data used in power system is increasing rapidly. In order to improve the efficiency of power …

Labels Need Prompts Too: Mask Matching for Natural Language Understanding Tasks

2024 · Proceedings of the AAAI Conference on Artificial Intelligence

Textual label names (descriptions) are typically semantically rich in many natural language understanding (NLU) tasks. In this paper, we incorporate the prompting methodology, which is widely used to enrich model input, into the label side …

BeNet: BERT Doc-Label Attention Network for Multi-Label Text Classification

2024 · Applied and Computational Engineering

Multi-label Text Classification (MLTC) holds significant importance and serves as a foundational aspect in Natural Language Processing (NLP), which aims at assigning multiple labels for a given document. Many real-world tasks can be viewed as …

Improved Bounds for Pure Private Agnostic Learning: Item-Level and User-Level Privacy

2024 · arXiv (Cornell University)

Machine Learning has made remarkable progress in a wide range of fields. In many scenarios, learning is performed on datasets involving sensitive information, in which privacy protection is essential for learning algorithms. In this work, …

Tackling Feature Skew in Heterogeneous Federated Learning with Semantic Enhancement

2024

A critical challenge in federated learning is data heterogeneity, compounded by varying local data distribution, thereby significantly impacting the performance of both local and global models. Prior works struggle to effectively address data heterogeneity but …

Underestimated Privacy Risks for Minority Populations in Large Language Model Unlearning

2024 · arXiv (Cornell University)

Large Language Models (LLMs) embed sensitive, human-generated data, prompting the need for unlearning methods. Although certified unlearning offers strong privacy guarantees, its restrictive assumptions make it unsuitable for LLMs, giving rise to various heuristic approaches …

Runtime-Aware Pipeline for Vertical Federated Learning with Bounded Model Staleness

2025

Vertical federated learning (VFL) enables a privacy-preserving collaboration among various parties to train a global model by melding their geo-distributed data features. Communication has been recognized as the primary bottleneck that impairs training efficiency due …

A Survey of Task-Oriented Knowledge Graph Reasoning: Status, Applications, and Prospects

2025

Knowledge graphs (KGs) have emerged as a powerful paradigm for structuring and leveraging diverse real-world knowledge, which serve as a fundamental technology for enabling cognitive intelligence systems with advanced understanding and reasoning capabilities. Knowledge graph …

Efficient shortest distance approximate query on large scale encrypted graph data

2025 · Cybersecurity

Abstract The problem of querying shortest distance on a graph has attracted significant research attention due to the widespread applicability of graphs and the ability of graph shortest path queries to address numerous application problems. …

Estimation of photovoltaic power generation in traditional protected villages in mountainous areas based on satellite image semantic segmentation and 3D terrain reconstruction

2025 · Journal of Asian Architecture and Building Engineering

With the growing demand for renewable energy, rooftop PV systems have gained widespread attention and adoption. However, existing methods for assessing PV potential are designed for urban areas and cannot address traditional protected villages or …

Novel Multi-Target Tracking Method: PMBM Filter Combined SVD-SCKF with GP-Driven Measurements

2026 · Sensors

Owing to multi-target tracking in scenarios with nonlinearity, uncertain measurement model and high clutter density, the Poisson multi-Bernoulli mixture (PMBM) recursion is prone to unstable covariance propagation under nonlinear dynamics as well as uncertainty in …

Exploration and Exploitation: A Study on Sample Efficiency in Reinforcement Learning With Multifaceted Curiosity Rewards and Adaptive Experience Replay Utilisation in Sparse Reward Environments

2026 · CAAI Transactions on Intelligence Technology

ABSTRACT In recent years, the widespread application of deep reinforcement learning (DRL) in autonomous systems has highlighted the importance of achieving high sample efficiency under sparse reward conditions. To improve sample efficiency in sparse reward …

Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling

2019 · arXiv (Cornell University)

Lingvo is a Tensorflow framework offering a complete solution for collaborative deep learning research, with a particular focus towards sequence-to-sequence models. Lingvo models are composed of modular building blocks that are flexible and easily extensible, …

Bytes Are All You Need: End-to-end Multilingual Speech Recognition and Synthesis with Bytes

2019

We present two end-to-end models: Audio-to-Byte (A2B) and Byte-to-Audio (B2A), for multilingual speech recognition and synthesis. Prior work has predominantly used characters, sub-words or words as the unit of choice to model text. These units …

Multilingual Speech Recognition with a Single End-to-End Model

2018

Training a conventional automatic speech recognition (ASR) system to support multiple languages is challenging because the sub-word unit, lexicon and word inventories are typically language specific. In contrast, sequence-to-sequence models are well suited for multilingual …