arXiv subjects

Liwen Ouyang

Publications and source records attributed to Liwen Ouyang.

5 recordsLinked to original sources

arXiv2026

Batched Paillier-Based Hamming-Distance Computation over Binary Embeddings

Additively homomorphic encryption supports outsourced computation on encrypted binary embeddings, but large-integer arithmetic and data movement can limit throughput. We describe a Paillier-based client that combines a carry-separated binary encoding, table-based encryption, reduced-exponent decryption, CUDA/CGBN arithmetic, persistent device state, and batched retrieval integration. We establish the encoding's correctness and characterize four CPU and GPU client configurations. The lookup configuration uses a 280-bit exponent-size parameter. Across 3 warm-state trials on batches of 10,000 random 512-bit embeddings, the lookup GPU configuration achieved median-batch throughputs of 43,091 encryptions/s and 28,983 Hamming- distance decodes/s. Its amortized costs were 0.0232 ms and 0.0345 ms per vector, corresponding to factors of 453.8 and 200.9 relative to the measured CPU baseline. These implementation- specific results demonstrate the throughput benefits of combining cryptographic precomputation, batched accelerator execution, and persistent runtime state. The study distinguishes warm-batch performance from isolated-request latency and identifies the remaining costs of initialization, transport, and retrieval integration.

cs.CR↗

arXiv2023

Integrating Offline Reinforcement Learning with Transformers for Sequential Recommendation

We consider the problem of sequential recommendation, where the current recommendation is made based on past interactions. This recommendation task requires efficient processing of the sequential data and aims to provide recommendations that maximize the long-term reward. To this end, we train a farsighted recommender by using an offline RL algorithm with the policy network in our model architecture that has been initialized from a pre-trained transformer model. The pre-trained model leverages the superb ability of the transformer to process sequential information. Compared to prior works that rely on online interaction via simulation, we focus on implementing a fully offline RL framework that is able to converge in a fast and stable way. Through extensive experiments on public datasets, we show that our method is robust across various recommendation regimes, including e-commerce and movie suggestions. Compared to state-of-the-art supervised learning algorithms, our algorithm yields recommendations of higher quality, demonstrating the clear advantage of combining RL and transformers.

cs.IR↗

arXiv2022

Maximum Mean Discrepancy for Generalization in the Presence of Distribution and Missingness Shift

Covariate shifts are a common problem in predictive modeling on real-world problems. This paper proposes addressing the covariate shift problem by minimizing Maximum Mean Discrepancy (MMD) statistics between the training and test sets in either feature input space, feature representation space, or both. We designed three techniques that we call MMD Representation, MMD Mask, and MMD Hybrid to deal with the scenarios where only a distribution shift exists, only a missingness shift exists, or both types of shift exist, respectively. We find that integrating an MMD loss component helps models use the best features for generalization and avoid dangerous extrapolation as much as possible for each test sample. Models treated with this MMD approach show better performance, calibration, and extrapolation on the test set.

cs.LG↗

arXiv2021

Estimation of Corporate Greenhouse Gas Emissions via Machine Learning

As an important step to fulfill the Paris Agreement and achieve net-zero emissions by 2050, the European Commission adopted the most ambitious package of climate impact measures in April 2021 to improve the flow of capital towards sustainable activities. For these and other international measures to be successful, reliable data is key. The ability to see the carbon footprint of companies around the world will be critical for investors to comply with the measures. However, with only a small portion of companies volunteering to disclose their greenhouse gas (GHG) emissions, it is nearly impossible for investors to align their investment strategies with the measures. By training a machine learning model on disclosed GHG emissions, we are able to estimate the emissions of other companies globally who do not disclose their emissions. In this paper, we show that our model provides accurate estimates of corporate GHG emissions to investors such that they are able to align their investments with the regulatory measures and achieve net-zero goals.

cs.LG↗

arXiv2015

Designed Sampling from Large Databases for Controlled Trials

The increasing prevalence of rich sources of data and the availability of electronic medical record databases and electronic registries opens tremendous opportunities for enhancing medical research. For example, controlled trials are ubiquitously used to investigate the effect of a medical treatment, perhaps dependent on a set of patient covariates, and traditional approaches have relied primarily on randomized patient sampling and allocation to treatment and control group. However, when covariate data for a large cohort group of patients have already been collected and are available in a database, one can potentially design a treatment/control sample and allocation that provides far better estimates of the covariate-dependent effects of the treatment. In this paper, we develop a new approach that uses optimal design of experiments (DOE) concepts to accomplish this objective. The approach selects the patients for the treatment and control samples upfront, based on their covariate values, in a manner that optimizes the information content in the data. For the optimal sample selection, we develop simple guidelines and an optimization algorithm that provides solutions that are substantially better than random sampling. Moreover, our approach causes no sampling bias in the estimated effects, for the same reason that DOE principles do not bias estimated effects. We test our method with a simulation study based on a testbed data set containing information on the effect of statins on low-density lipoprotein (LDL) cholesterol.

stat.ME↗