The ranking stage serves as the central optimization and allocation hub in advertising systems, governing economic value distribution through eCPM and orchestrating the user-centric blending of organic and advertising content. Prevailing ranking models often rely on fragmented modules and hand-crafted features, limiting their ability to interpret complex user intent. This challenge is further amplified in location-based services such as food delivery, where user decisions are shaped by dynamic spatial, temporal, and individual contexts. To address these limitations, we propose a novel generative framework that reframes ranking as a context comprehension task, modeling heterogeneous signals in a unified architecture. Our architecture consists of two core components: the Generative Contextual Encoder (GCE) and the Generative Contextual Fusion (GCF). The GCE comprises three specialized modules: a Personalized Context Enhancer (PCE) for user-specific modeling, a Collective Context Enhancer (CCE) for group-level patterns, and a Dynamic Context Enhancer (DCE) for real-time situational adaptation. The GCF module then seamlessly integrates these contextual representations through low-rank adaptation. Extensive experiments confirm that our method achieves significant gains in critical business metrics, including click-through rate and platform revenue. We have successfully deployed our method on a large-scale food delivery advertising platform, demonstrating its substantial practical impact. This work pioneers a new perspective on generative recommendation and highlights its practical potential in industrial advertising systems.

Recently, the topic of table pre-training has attracted considerable research interest.However, how to employ table pre-training to boost the performance of tabular prediction remains an open challenge.In this paper, we propose TAPTAP, the first attempt that leverages table pre-training to empower models for tabular prediction.After pre-training on a large corpus of real-world tabular data, TAPTAP can generate high-quality synthetic tables to support various applications on tabular data, including privacy protection, low resource regime, missing value imputation, and imbalanced classification.Extensive experiments on 12 datasets demonstrate that TAPTAP outperforms a total of 16 baselines in different scenarios.Meanwhile, it can be easily combined with various backbone models, including LightGBM, Multilayer Perceptron (MLP) and Transformer.Moreover, with the aid of table pre-training, models trained using synthetic data generated by TAPTAP can even compete with models using the original dataset on half of the experimental datasets, marking a milestone in the development of synthetic tabular data generation.The code and datasets are available at https: //github.com/ZhangTP1996/TapTap.

The goal of automated feature generation is to liberate machine learning experts from the laborious task of manual feature generation, which is crucial for improving the learning performance of tabular data. The major challenge in automated feature generation is to efficiently and accurately identify effective features from a vast pool of candidate features. In this paper, we present OpenFE, an automated feature generation tool that provides competitive results against machine learning experts. OpenFE achieves high efficiency and accuracy with two components: 1) a novel feature boosting method for accurately evaluating the incremental performance of candidate features and 2) a two-stage pruning algorithm that performs feature pruning in a coarse-to-fine manner. Extensive experiments on ten benchmark datasets show that OpenFE outperforms existing baseline methods by a large margin. We further evaluate OpenFE in two Kaggle competitions with thousands of data science teams participating. In the two competitions, features generated by OpenFE with a simple baseline model can beat 99.3% and 99.6% data science teams respectively. In addition to the empirical results, we provide a theoretical perspective to show that feature generation can be beneficial in a simple yet representative setting.

Yanan Zheng, Jing Zhou, Yujie Qian, Ming Ding, Chonghua Liao, Li Jian, Ruslan Salakhutdinov, Jie Tang, Sebastian Ruder, Zhilin Yang. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

Most previous methods for text data augmentation are limited to simple tasks and weak baselines.We explore data augmentation on hard tasks (i.e., few-shot natural language understanding) and strong baselines (i.e., pretrained models with over one billion parameters).Under this setting, we reproduced a large number of previous augmentation methods and found that these methods bring marginal gains at best and sometimes degrade the performance much.To address this challenge, we propose a novel data augmentation method FlipDA that jointly uses a generative model and a classifier to generate label-flipped data.Central to the idea of FlipDA is the discovery that generating labelflipped data is more crucial to the performance than generating label-preserved data.Experiments show that FlipDA achieves a good tradeoff between effectiveness and robustness-it substantially improves many tasks while not negatively affecting the others. 1

In this paper, we proposed a path planning algorithm based on Q-learning model to simulate an environment model, which is suitable for the complex environment. A virtual simulation platform has been built to complete the experiments. The experimental results show that the algorithm proposed in this paper can be effectively applied to the solution of vehicle routing problems in the complex environment.

Aiming at the problem that single‐network hopping signals have not fully utilised its frequency domain sparse characteristic in the parameter estimation, this study proposes a parameter estimation of frequency hopping (FH) signal based on multi‐measurement vector sparse Bayesian learning (MSBL) in modulation wideband converter (MWC). Since the FH signal is sparse in the frequency domain, the authors apply the MSBL method to estimate its parameters. After the signal is sampled by the MWC, the MSBL algorithm is used to reconstruct its support set. Then the time–frequency ridge method is used to estimate the signal's hop duration, time‐hopping, and carrier frequency based on the time–frequency map. Simulation experiments show that under the condition of low signal‐to‐noise ratio, the parameter estimation performance in the case can be improved by up to 65% and anti‐noise performance can be improved up to 6 db compared with the existing method. The result is very close to the Nyquist full sampling and can greatly improve the accuracy of the FH signal parameter estimation in the MWC system and relieve the pressure of the hardware.

This paper presents the design, analysis, and experimental validation of a fault detection and identification (FDI) scheme for dc-dc power electronic converters. The FDI scheme includes two new classes of fault filters: (1) a linear-switched fault detection (FD) filter and (2) a bank of linear-switched fault identification (FI) filters. Both the FD filter and the bank of FI filters have a structure similar to that of Luenberger observer. The FD filter detects a fault event and the bank of FI filters identify a faulted converter component, for instance, failure in switching or passive components. We present simulation and experimental results for a prototype 2 kW solar photovoltaic (PV) boost dc-dc converter system to demonstrate the efficacy of the proposed FDI scheme. The experimental results demonstrate accurate fault detection and identification for a collection of catastrophic component faults in PV dc-dc power converters.

The genetic particle swarm optimization (GPSO) was derived from the original particle swarm optimization (PSO), which is incorporated with the genetic reproduction mechanisms, namely crossover and mutation. To solve traveling salesman problems (TSP), a modified genetic particle swarm optimization (MGPSO) was introduced, where the new solution was generated with local best and individual best solutions with crossover and mutation operators. MGPSO was implemented to the well-known TSP and by comparison with the results of the original PSO, MGPSO has provided much better performance. Furthermore, MGPSO was employed to solve TSP with time windows, where besides minimizing the route, the truck were required to arrive at specifically during a time window, which made the TSP to be a constrained combinatorial optimization. To solve the constraints, the stochastic ranking algorithm was introduced The approach was experimented with the well-known TSP case. The simulation results have shown its robust and consistent effectiveness.

The resource planning and scheduling technology of payload is a key technology to realize an automated control for earth observing satellite with limited resources on satellite, which is implemented to arrange the works states of various payloads to carry out missions by optimizing the scheme of the resources. The scheduling task is a difficult constraint optimization problem with various and mutative requests and constraints. Based on the analysis of the satellite's functions and the payload's resource constraints, a proactive planning and scheduling strategy based on the availability of consumable and replenishable resources in time-order is introduced along with dividing the planning and scheduling period to several pieces, where then the planning and scheduling is modeled as a combinatorial optimization. The genetic particle swarm optimization algorithm (GPSO) is proposed to address the problem, which was derived from the original continuous particle swarm optimization (PSO) and incorporated with the genetic reproduction mechanisms, namely crossover and mutation. The simulation results have shown that GPSO significantly improved the search efficacy of PSO for the combinatorial optimizations.

This paper proposes a new information filtering method which is used to bridge academic research and business interests. Sometimes terms which are used in business situations have different meanings from those of academic research. Hence, although business persons want to find academic research which matches the needs of their business interests, search results usually contain many unintended material. In order to remove such unintended results, we propose a new filtering method which uses related documents as reference material.

In this paper, we introduce a speaker depedent isolated word recognition system which is dedicated for Chinese character input. The method presented here offers an effective solution to the large-vocabulary isolated word recognition. The classification is performed hierarchically. In the preselection stage, the candidates are selected by KNN rule using the utterance length and coarse templates. A ranked list of candidates is obtained by the following final classifier which performs a nonlinear time alignment of uncompressed templates using DTW algorithm. The recognition rate is more than 90% when monosyllables take up 1/3 of the vocabulary (1000).