Papers

Papers Digital, Intelligence & Platforms

Neural Combinatorial Optimization for Vehicle Routing: A 4-Category Framework

This paper provides a comprehensive survey of Neural Combinatorial Optimization (NCO) algorithms for solving Vehicle Routing Problems (VRPs), categorizing them into four distinct types: Learning to Construct (L2C), Learning to Improve (L2I), Learning to Predict-Once (L2P-O), and Learning to Predict-Multiplicity (L2P-M). The study highlights the current limitations and future directions in NCO, aiming to enhance the efficiency and scalability of VRP solutions.

2026-09-23

Papers Digital, Intelligence & Platforms

Transformer Achieves 28.4 BLEU on WMT 2014 English-to-German Translation

The Transformer, a novel neural network architecture introduced by Ashish Vaswani, Noam Shazeer, and colleagues, achieves state-of-the-art results in machine translation tasks. By relying solely on self-attention mechanisms, the Transformer outperforms traditional recurrent and convolutional models, offering superior performance, faster training, and better parallelization.

2026-09-20

Papers Digital, Intelligence & Platforms

Enhancing Recommender Systems with Graph Neural Networks: A Comprehensive Survey

Recommender systems (RS) are essential for navigating the vast array of products and services online. Traditional RS, such as content-based and collaborative-filtering, struggle with complex, non-Euclidean data like Knowledge Graphs (KG). This survey by Gao et al. provides a detailed taxonomy of GNN-based Knowledge Aware Deep Recommender (GNN-KADR) systems, highlighting their effectiveness in addressing practical recommendation issues.

2026-09-19

Papers Digital, Intelligence & Platforms

Sample-Efficient Reinforcement Learning via Counterfactual Data Augmentation

Chaochao Lu, Biwei Huang, et al. propose a sample-efficient reinforcement learning (RL) algorithm that leverages structural causal models (SCMs) and counterfactual reasoning to address data scarcity and mechanism heterogeneity. The method enhances policy learning in scenarios with limited data, such as healthcare, by generating augmented datasets for more reliable and personalized policies.

2026-08-29

Papers Digital, Intelligence & Platforms

AdaRL: Efficient and Reliable Policy Adaptation with 50-70 Samples in Transfer Reinforcement Learning

AdaRL, proposed by Biwei Huang, Fan Feng et al., introduces a framework for efficient and reliable policy adaptation in transfer reinforcement learning. By leveraging a parsimonious graphical representation, AdaRL can adapt policies to new environments with minimal data, achieving high success rates in Cartpole and Atari games.

2026-08-29

Papers Digital, Intelligence & Platforms

Enhancing RL Generalization with Compositional Causal Components

The paper "Modeling Unseen Environments with Language-guided Composable Causal Components in Reinforcement Learning" by Xinyue Wang and Biwei Huang introduces a novel framework, WM3C, that enhances reinforcement learning (RL) generalization by decomposing tasks into composable causal components. This approach leverages language to guide the decomposition of the latent space, leading to better generalization in unseen environments.

2026-08-12

Papers Digital, Intelligence & Platforms

Algorithmic Labor and Temporality in Chinese Food Delivery: A Study of Platform Workers

Ping Sun's study explores the labor conditions of delivery workers on food delivery platforms in China, focusing on how they experience and make sense of algorithms. The research reveals that these workers are not passive entities but active participants in creating their own "organic algorithms" to manage and sometimes subvert the system.

2026-08-12

Papers Digital, Intelligence & Platforms

Representation Learning Advances: 30% Improvement in Speech Recognition

The paper by Yoshua Bengio, Aaron Courville, and Pascal Vincent reviews recent advancements in representation learning and deep learning. It highlights significant improvements in speech recognition, natural language processing, and object recognition, demonstrating the potential of these techniques to revolutionize AI and supply chain applications.

2026-08-12

Papers Digital, Intelligence & Platforms

Meituan Cuts Order Cancellations by 25% with AI Bonus Framework

Meituan’s AI-powered Multi-Stage Bonus Allocation framework cut daily order cancellations by over 25%, reducing 165,000 NA-canceled orders and cutting restaurant food waste compensation by 30%. Built with a semi-black-box model and Lagrangian dual-based dynamic programming, it makes millisecond bonus decisions across order lifecycle stages. Deployed at scale, it demonstrates how real-time incentive optimization strengthens last-mile supply chain resilience — a lesson increasingly relevant for e-commerce logistics, perishable goods distribution, and time-sensitive service platforms globally.

2026-03-30

Papers Digital, Intelligence & Platforms

**A Framework for Multi-Stage Bonus Allocation in Meal Delivery Platforms: Operationalizing Real-Time Incentive Optimization at Scale**

*SCI.AI Research Section — In-Depth Technical & Business Analysis* --- ### **Research Background: The Hidden Crisis of Order Cancellation in On-Demand Food Delivery** Food delivery platforms operat...

2026-03-26

Papers Digital, Intelligence & Platforms

A Multi-stage Bonus Allocation Framework for Meal Delivery Platforms

This research proposes a multi-stage bonus allocation framework to address order acceptance challenges in meal delivery platforms. Through dynamic bonus incentives, the framework can reduce canceled orders by over 60% within a limited budget while saving restaurants more than 30% in food waste compensation.

2026-03-17

Papers Digital, Intelligence & Platforms

How Meituan Reduces Order Cancellations by 25% with Dynamic Bonus Allocation: KDD 2022 Research from Huazhong University

Meituan and Huazhong University research presents Multi-Stage Bonus Allocation (MSBA) framework, optimizing subsidy fund efficiency through Lagrangian dual dynamic programming, achieving 25% order cancellation reduction and 31% restaurant compensation cost savings in real-world A/B testing. This research demonstrates how operations research and machine learning can be combined to solve real-world logistics optimization problems at massive scale.

2026-02-27