pandascute's Projects
2018-CCF大数据与计算智能大赛-面向电信行业存量用户的智能套餐个性化匹配模型联通赛-复赛第二名-【多分类,embedding】
2018年甜橙金融杯大数据建模大赛-初赛第四-复赛线上11-决赛9-复现top1解决方案-【二分类,风控】
Codes and dashboards for 4th place solution for Kaggle's Home Credit Default Risk competition
人工智能学习路线图,整理近200个实战案例与项目,免费提供配套教材,零基础入门,就业实战!包括:Python,数学,机器学习,数据分析,深度学习,计算机视觉,自然语言处理等热门领域
AI Challenger 2018 细粒度用户评论情感分析,排名17th,基于Aspect Level 思路的解决方案
2018/2019/校招/春招/秋招/算法/机器学习(Machine Learning)/深度学习(Deep Learning)/自然语言处理(NLP)/C/C++/Python/面试笔记
:blue_book:分享最新BAT面试题(包含机器学习,Linux,PHP,大数据,Python,Java,前端...),持续更新ing...
Preliminary rank_13
2017年CCF大数据与计算智能大赛-小超市供销存管理优化
BDCI 2018 汽车行业用户观点主题及情感识别 《Just a test》团队方案-决赛第一名
TensorFlow code and pre-trained models for BERT
Deep Learning 101 with PaddlePaddle (深度学习框架入门教程)
Playing with the IEEE-CIS Vesta fraud detection competition
The code for CCF-BDCI-Sentiment-Analysis-Baseline
汽车行业用户观点主题及情感识别
ccf 2020 qa match competition top1
简单的实现推荐系统的召回模型和排序模型,其中召回模型使用协同过滤算法,排序模型使用gbdt+lr算法
Crystal Graph Neural Networks
参加过的比赛记录,包括代码和经验和其他文件
KDD CUP 2019 Baseline
消费者人群画像—信用智能评分比赛开源
CS224n: Natural Language Processing with Deep Learning Assignments Winter, 2017
CTR prediction using FM FFM and DeepFM
Data competition Top Solution 数据竞赛top解决方案开源整理
Traditional approaches to production equipment maintenance remain mainly in corrective maintenance and routine-based maintenance. Corrective maintenance, which is no fail no repair, implies unscheduled downtime of equipment, severe security risk and unacceptable economic loss, while routine or time-based maintenance is time consuming and costly. By data analysis and modeling, predictive maintenance forecasts the residual life or possible failure of key components, and schedules maintenance accordingly,in order to avoid unscheduled downtime and to save cost.
We extract actual SCADA data of a certain offshore wind farm and artificially remove some of the data, including but not limited to deleting all data of a certain time period, data of a certain turbine at a certain time, some fields of a certain turbine at a certain time, etc. The applicant needs to use the remaining data to recover the deleted data, and the accuracy of the recovery shall be the final evaluation benchmark
Easy-to-use,Modular and Extendible package of deep-learning based CTR models.DeepFM,DeepInterestNetwork(DIN),DeepCrossNetwork(DCN),AttentionalFactorizationMachine(AFM),Neural Factorization Machine(NFM),AutoInt
face recognition application
Apache Flink