.png)
HOCIntelligentTechnologyGroup
Faculty, BIG DATA (information), Large language model(LLM), Generative Pre-trained Transforme model(GPT), Research Lab's "FBLGR Quality" is a leading company,AI Science and Technology Innovation for Global Sustainable Development.



Research ScienceTeam
动机驱动、基于热情、拼命学习,学无止境
大胆地走以前没有人走过的路、大胆做没人做过的事



哈牛桥智能科技南京有限公司(简称:哈牛桥智能科技)成立于2019年7月。注册资本金5000万,哈牛桥智能科技是依托哈佛大学,牛津大学,剑桥大学等为主的名校的研究学术界的领域领先顶尖博士教授研发团队和强大的技术积累,超过400人 ,大数据解决方案以行业“智能化、数字化”转型为 契机,深耕“金融、交通、运营商” 三个垂直领域,提供专业的智能数 据服务,总部南京,在上海、北京、 深圳、杭州等地设有项目交付中心,哈牛桥智能科技南京公司由国龙剑桥博士于红红哈佛博士领导的牛津大学剑桥大学哈佛大学MIT等博士团队,ERP咨询,云技术,大数据,区块链,AI机器学习,RPA, OCR-AI 特别是深度学习等人工智能领域实践能力专家,40年工作经验专业团队。哈牛桥智能科技是中国国家级 A类团队的,也是得到国家科技扶持金的科技家团队中国国家的重点科研室中国科学院的院士,清华大学的教授。和技术大趋势的融合,领先的分析技术被用于政府/地方政府,教育/医疗/医疗保健,金融,制造业,物流,通讯/广播,建筑/房地产,电力/燃气/水,网络,制药,农业,零售,制造,交通,体育,宇宙航空,广告,IOT,ICT等行业。中国科学院院士,IEEE会员,ERP区块链云技术大数据人工智能相关领域(不限于语音处理领域,包含人工智能各领域)的专利产品大赛获得一等奖了,并和阿里,腾讯,华为等多家签约达成。AI×5G 刷脸成为主流支付方式,所见即所买、短视频AI動画×5G、Metaverse,智能驾驶,金融,5G 远程医疗 5G AI 医疗,智能驾驶 智能商业 智能医疗公安 5G机器人 5G材料半导体 体育 娱乐等领域技术是我们的主流、在分离纯化、创新药物、生物技术、芯片设计、量子点显示、多点触控、纳米微球和低碳纳米材料等关键技术世界领先 智能驾驶,智能制造,机器人,智能医疗。人脸与人体分析技术,SLAM 与 3D 视觉,通用与专业图像识别,机器人控制与传感,海量视频理解与挖掘,图像视频处理增强医学图像分析,人工智能计算平台,AI 超算平台,自研训练框架,AI 高性能存储 高性能异构计算
主营业务在美国英国日本- 系统和移动应用程序开发。
南京(中国) - 中国科技研究的专业服务 信息科学、生物科学和材料科学的系统开发。
• 深圳(中国)——为中国企业开发环境和能源系统。
• 上海(中国)——中国的自动驾驶,医疗公司的系统开发。
• 杭州(中国)——为中国制造公司DX 进行系统开发。
• 无锡(中国) - 中国公司的离岸开发、SI 和 BPO。






China
南京总部
上海
北京
杭州
无锡
〒210089
中国南京市江北新区星火路17号创智大厦801室-845
TEL:15358411774
*作为防止新型冠状病毒传播的措施的一部分,我们将暂时停止接受电话咨询。
email:1500467240@qq.com

United States of America
New York WashingtonD.C.
california new jersey
England London
Germany BERLIN
Canada Toronto
Japan osaka Tokyo Nagoya







Story





以原创技术体系为根基,深度学习平台为核心“大脑”,布局多领域、多方向前沿研究,快速打通AI在各个垂直场景中的应用。
研究始终坚持原创,不断突破创新,并已与国内外数十所一流高校和科研院所建立深度合作关系,具有深厚的人工智能领域学术积累和健全的人才培养机制。The Harvard Graduate School of Design 哈佛大学设计学院,被认为是世界上最好的设计学院之一合作,哈佛大学,剑桥大学牛津大学MIT,东京大学京都大学大阪大学,清华大学,中国科学院等名校合作
资质荣誉
国家级高新技术企业 •南京市双软企业 •江苏省重点民营高科技企业 •CMMI3 •ISO9001 •ISO27001
研发中心南京总部,杭州总部,上海总部,台州分部,丽水分部,苏州分部等,全球 日本,英国,美国,韩国。。。
核心优势
哈牛桥智能科技是
拥有硅谷优秀的机器学习建模算法,我们对机器学习和深度学习在科技领域的应用展开不断探索与追求。目前在机器学习方面的技术储备(指我司模型团队自主开发的核心算法的代码)包括:
-
随机森林,神经网络,支持向量机,决策树类的单个模型以及上述单个模型的bagging算法
-
适用于机器学习的cross validation算法
-
各种基础算法的stacking方法实现
-
决策树的gradient boosting machine 算法
-
基于SVM的半监督学习等。
-
主要面向软件技术、人工智能等领域开展基础研究和应用研究、应用示范及产业化,致力于突破核心关键技术,加速成果转移转化,建设软件产业新型智库,打造软件产业人才高地。共同研发合作伙伴是中科院软件所和剑桥大学 公司在软件领域的技术和人才优势,公司自2019年7月成立以来快速成长,现已布局智能系统软件、服务机器人通用软件平台、智能无人系统训练与评估平台、云计算与大数据、物联网城市大脑等五个具体科研方向.农业、能源、电子计算、激光、空间科学、高能物理和通用工程8个领域的高技术综合研究和自然科学顶尖研究参与了多个大型项目。 火箭、人造卫星 以光电技术为特色,集光学、力学、电学、计算机、材料为特色,与多部门合作发展的研究型企业,中科院参与附属光学精密仪器与物理研究所和神舟载人飞船的共同研究。
剑桥大学共同研发医药领域的开发
-
中国深圳创新创业大赛国际大赛中哈牛桥智能科技南京有限公司获得决赛奖
-
参赛项目全球最大区块链人工智能大数据云平台,专注打造“研发”全球
National Institute of Information and Communications Technology NICT
Future Engineering Research Institute
Beyond 5G R&D workshop research presentation
Research Team of Japan HOC Intelligent Technology Co., Ltd.
Research presentation by YuHonghong, Representative Director, and Guolong, Director
(1) AI processing x quantum computing technology Powerful computing capacity x Beyond 5G communication service
(2) Brain-inspired AI x quantum ICT, security technology, etc.) ・Quantum cryptographic communication


















合作伙伴認証資格














.png)












.png)

















Team members




Client




















中非共建“一带一路”合作,Second-hand clothes are sold to Africa.In the African second-hand clothing market, China has become the largest exporter in Africa for the first time
Every year, everyone buys new clothes, and in 2018, 54.06 billion pieces of clothing were sold nationwide.
But few people pay attention to old clothes, new clothes buy and buy, so how to deal with old clothes?
01 Never too many clothes
In rural areas, the general lifespan of clothes is 3-4 years, especially for many middle-aged and elderly people, and they are reluctant to throw them away when they are old.
In towns, the general lifespan of clothes is 1-2 years, and for young people the use time is even shorter. But many old clothes were either discarded or piled up in the closet.
According to idle fish data, the idle old clothes produced by ordinary Chinese families weigh 60 pounds a year.
In today's era of women's clothing consumption, women buy clothes at least twice a month, and trendy clothing only stays for one season. 80% of these clothes are forgotten after the season, but the number of clothes purchased is still increasing at a rate of 15% per year.
Men have a lower impulse to buy clothes, purchasing on average once a month. But they don't give away their old clothes, and they rarely throw away their old clothes.
02 Lost to half of the Daqing oilfield in one year
The data shows that the average person in China buys about 10 pieces of new clothes per year, of which 3-5 are discarded. Between 26 million and 28 million tons of clothes are discarded every year, with a utilization rate of less than 1%, which is a huge waste.
Some people know that an important raw material for clothing fiber is crude oil, which is equivalent to discarding 24 million tons of petroleum crude oil every year, more than half of the Daqing oilfield production.
So at present, the treatment of old clothes is mainly the following 4 ways:
A. Public welfare donations
There will be many non-profit organizations or institutions online that will collect used clothes and regularly send them to poor mountainous areas in central and western China, or donate them to local people in need through charities.
In addition, in many first- and second-tier cities, there are some used clothes recycling bins. Some of these recycling bins are jointly organized by the government and enterprises, and some of the clothes will be sorted and used for charitable donations.
However, due to information asymmetry, there are now many public welfare organizations or poor schools, and old clothes are almost impossible to put down.
The first batch has not yet been processed, and the second batch has come again. Some departments received donations in 2009 and still have a backlog of hundreds of thousands of used clothes.
And the cost of used clothes donation is actually very high, from recycling, sorting, disinfection, cleaning, packaging, distribution and other links, the donation cost of an old clothes is about 12-15 yuan.
B. Sell or give away
More than 500,000 pieces of used clothes are sold every month, of which post-95 young people are the main force, accounting for 34%.
Post-95 young people are not only willing to buy old clothes, but also accept old clothes donations, and 30% of the used clothes donations in Idle Fish are included in the post-95s.
Of course, women are more willing to deal with second-hand clothing sales, 4 times more than men
C. Reprocessing
Whether it is an individual vendor or an organization, eventually the vast majority of used clothes are recycled and sorted by recycling plants.
Used clothes that cannot be sold twice, but that are not used for burning.
After sorting, it will be sold to downstream suppliers for recycling, and made into greenhouse insulation cotton, environmentally friendly gloves, environmentally friendly bags, sound insulation cotton, building materials, etc.
D. Resale
Either way, it is not the most impressive in terms of economic benefits.
For businesses specializing in the recycling of used clothes, it is only by sorting out about 90% of new clothes and selling them for secondary sales to make money.
At present, the living conditions of most families in the country have been significantly improved, and most of them will not buy old clothes to wear.
Because in China, it is very convenient to buy clothes, very cheap, there are many styles, and a very small number of used clothes flow into the local market.
So where are these huge quantities of old clothes being sold twice?
The answer is: Africa! Yes, it's Africa, and it's very profitable!
Africa"The World's Largest E-Waste Graveyard" Agbogbloshie, Ghana
From a medium- to long-term perspective, we will adopt cutting-edge technologies and proactively promote concrete measures that contribute to social structural transformation toward a decarbonized and recycling-oriented society that is expected to grow sustainably in the future.
300 million mobile phones discarded in Africa annually "Leaked from Japan" As for the recycling of e-waste, "there is a connection between Africa and Europe, and an international chain of resource circulation has been established. It can be expanded to other resources as well. I think the environmental value behind the business is great
In addition to promoting cross-border e-commerce (EC) as part of China's "digital innovation project" jointly implemented with Africa, live commerce events are also being held at the same time to introduce excellent African-made products such as coffee and crafts.
人才概览/ Overview
科学研究实验室全力构建开放创新的人才生态,目前人才队伍总体规模已突破4000人,其中全职员工达2600余人, 学术带头人近300人,研究人员中博士以上学历占比超99%,组建了一支“领域专精、层次高端、梯队有序”的人才队伍。 实验室秉持分层分类、实战育才的人才培养理念,依托“科学研究书院”底座统筹建设人才培育体系,重点打造“领航计划”等培养项目。 实验室坚持以创新质量和实际贡献为人才评价核心导向,贯通人才发展各项通道,全方位激发人才创造活力, 为打造国家战略科技力量提供强有力的人才支撑。
院士团队 / Academician Team
由国内外著名专家组成,进行学术咨询和指导,负责向理事会提供有关创新研究方向、重点发展领域 、重大研究任务和目标等学术问题的咨询意见和建议,指导和把握实验室科研方向,进行学术工作评估。
成功合作医院
江苏省人民医院江苏省中医院南京市鼓楼医院南京市中医院南京市第一医院东南大学附属中大医院南京市儿童医院南京医科大学第二附属医院南京脑科医院江苏省肿瘤医院浙江省人民医院浙江大学医学院附属妇产科医院浙江省中医院浙江大学医学院附属第一医院(浙江省第一医院)浙江大学医学院附属儿童医院(浙大儿童医院、浙江省儿童医院、浙江省儿童保健院)浙江医科大学附属第一医院等

.jpg)



.png)










系统介绍
现有的基于AI的临床决策支持系统(CDSS)均是基于知识的,并且主要服务于诊断。另一方面,AlphaGo击败世界冠军李世石,使大家意识深度强化学习能利用过程数据发现新的围棋知识或者模式,从而可以超越人类。
临床决策支持系统,同样可以通过深度强化学习来大幅度提升智能水平,尤其疾病治疗方向。美国哈佛&MIT医学院等2019年在Nature Medicine发表论文,关于强化学习在医疗健康的技术应用指南。为采用强化学习进行患者治疗决策提供指导方针,他们希望这将加快观察群组以安全、有风险意识的方式告知医疗实践的速度。
我们的人工智能团队从2018年开始了这方面的研发工作,包括数据建设、算法开发、软件系统开发等。我们的产品取名凌波微步。凌波微步取自于《易经》,有精准的意思。
数据建设方面,获得了重症监护数据库MIT的访问资质,然后进行了本地化,开发出了数据对齐软件,以及基于因果检验的降维工具。算法开发方面,针对脓毒症的治疗策略,我们用Deep Deterministic Policy Gradient (DDPG)算法,超越了MIT团队的DQN算法,使患者获得更高的存活率。同时,模型训练时间更快,稳定性更好。软件系统开发方面,集成了数据和算法,实时获取来自HIS、LIS的数据,通过可视化界面定时输出每一位病人的疾病状态和治疗建议,供医生参考。
除了脓毒症的治疗策略,我们还分别和几个三甲医院开发呼吸衰竭、肾衰竭、肿瘤等疾病的治疗策略。
现已联合的博士后研究所有复旦大学、上海交通大学、浙江大学、华东师范大学、南京大学、中国科学院等学术科研机构。
——基于深度强化学习的临床决策支持系统
.png)







