Xinchi Qiu

Xinchi Qiu

Auto-generated profile, based on publicly available information

Xinchi Qiu

University of Cambridge, Cambridge, United Kingdom

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 May 2023
Verified
Conference Paper
Open access
CMChenyang Ma
Chenyang Ma
XQXinchi Qiu
Xinchi Qiu
DBDaniel Beutel
Daniel Beutel
NLNicholas Lane
Nicholas Lane
The privacy-sensitive nature of decentralized datasets and the robustness of eXtreme Gradient Boosting (XGBoost) on tabular data raise the needs to train XGBoost in the context of federated learning (FL). Existing works on federated XGBoost in the horizontal setting rely on the sharing of gradients, which induce per-node level communication frequency and serious privacy concerns. To alleviate these problems, we develop an innovative framework for horizontal federated XGBoost which does not depend on the sharing of gradients and simultaneously boosts privacy and communication efficiency by making the learning rates of the aggregated tree ensembles learnable. We conduct extensive evaluations on various classification and regression datasets, showing our approach achieves performance comparable to the state-of-the-art method and effectively improves communication efficiency by lowering both communication rounds and communication overhead by factors ranging from 25x to 700x.

24 August 2024
Verified
Conference Paper
Open access
TXTong Xia
Tong Xia
AGAbhirup Ghosh
Abhirup Ghosh
XQXinchi Qiu
Xinchi Qiu
CMCecilia Mascolo
Cecilia Mascolo
Federated Learning (FL) enables model development by leveraging data distributed across numerous edge devices without transferring local data to a central server. However, existing FL methods still face challenges when dealing with scarce and label-skewed data across devices, resulting in local model overfitting and drift, consequently hindering the performance of the global model. In response to these challenges, we propose a pioneering framework called FLea, incorporating the following key components: i) A global feature buffer that stores activation-target pairs shared from multiple clients to support local training. This design mitigates local model drift caused by the absence of certain classes; ii) A feature augmentation approach based on local and global activation mix-ups for local training. This strategy enlarges the training samples, thereby reducing the risk of local overfitting; iii) An obfuscation method to minimize the correlation between intermediate activations and the source data, enhancing the privacy of shared features. To verify the superiority of FLea, we conduct extensive experiments using a wide range of data modalities, simulating different levels of local data scarcity and label skew. The results demonstrate that FLea consistently outperforms state-of-the-art FL counterparts (among 13 of the experimented 18 settings, the improvement is over 5%) while concurrently mitigating the privacy vulnerabilities associated with shared features. Code is available at https://github.com/XTxiatong/FLea.git.

7 April 2021 in arXiv
Verified
Preprint
Open access
PDF file
AMAkhil Mathur
Akhil Mathur
DBDaniel J. Beutel
Daniel J. Beutel
PGPedro Porto Buarque de Gusmão
Pedro Porto Buarque de Gusmão
JFJavier Fernandez-Marques
Javier Fernandez-Marques
TTTaner Topal
Taner Topal
NLNicholas Lane
Nicholas Lane
Federated Learning (FL) allows edge devices to collaboratively learn a shared prediction model while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store data in the cloud. Despite the algorithmic advancements in FL, the support for on-device training of FL algorithms on edge devices remains poor. In this paper, we present an exploration of on-device FL on various smartphones and embedded devices using the Flower framework. We also evaluate the system costs of on-device FL and discuss how this quantification could be used to design more efficient FL algorithms.

6 June 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
WSWilliam F. Shen
William F. Shen
XQXinchi Qiu
Xinchi Qiu
NLNicholas D. Lane
Nicholas D. Lane
AIAlex Iacob Iacob
Alex Iacob Iacob
ZZZichen Zhang
Zichen Zhang
YZYulong Zhang
Yulong Zhang
Discovering effective alphas from noisy, non-stationary financial data remains a challenging open problem. Recent LLM-based alpha discovery systems improve search automation, but they still largely operate within human-specified research protocols and the evolving object is typically the alpha formulas or mining trajectory, rather than the researcher itself. To address these gaps, we introduce STAR, a self-tuning agent framework for studying alpha-research policy evolution. STAR couples an LLM-based alpha researcher with a metacognitive module that can revise the research scaffold, while fixed and isolated validation/test data and a hostside evaluator preserve the integrity of the utility channel, and the search policy prioritizes lineages with stronger descendants. Under a leakage-controlled forward protocol on the most recent 2026Q1 deployment window on CSI300, evolved STAR researchers generate substantially higher quality alphas than the base model and a comprehensive set of strong baselines. Qualitative analysis further indicates that these gains arise from genuine researcher evolution, including a 6.9× expansion of operator vocabulary, new tools, and improved research protocols. These results highlight the effectiveness of self-evolving agents in improving capabilities in noisy, weak-supervision domains, and position them as a promising pathway toward building more capable financial researchers.