About关于我 Skills技能世界 Projects项目经历 Internships实习经历 Contact联系我
HELLO, I'M你好,我是

Huang Yutong 黄雨彤

Welcome to my space! I'm an Economic Statistics major at Zhejiang Gongshang University, driven by curiosity and a passion for exploring the evolving intersection of AI and data.欢迎来到我的空间!我目前就读于浙江工商大学经济统计学专业,我充满好奇心,不断探索AI与数据融会贯通的新世界!

I thrive on diving into massive datasets and specializing in data analysis. My strength lies in combining statistical models with machine learning to uncover hidden value and deliver actionable solutions. While data analysis built my professional foundation, product design sharpened my logic, ensuring I approach every problem with a rigorous, analytical mindset.我热衷于深挖海量数据,深耕数据分析,擅长结合统计模型与机器学习来挖掘数据价值,输出可落地方案。数据分析培养了我的专业,产品设计让我具有更强的逻辑,让我始终秉持严谨的态度去解决每一个问题。

Outside of work and academics, you'll usually find me behind a camera, exploring the outdoors, or at the gym. I love capturing life's little joys, constantly challenging myself, and striving for continuous growth.在工作与学习之余,你通常会在镜头背后、或是户外和健身房里找到我。我热爱捕捉生活中的小乐趣,也不断挑战自我、持续成长。

I'm so glad you're here to share this journey with me!很高兴你能在这里,与我一同分享这段旅程!

Data Analysis数据分析
Product Design产品设计
Visual Creation视觉创作
Main Photo

Zhejiang Gongshang University浙江工商大学

Economic Statistics经济统计学

4.01 /5.0

I specialize in turning raw, messy data into clear, actionable insights using Python, R, and other statistical tools. Whether I'm cleaning datasets or building predictive models, my favorite part of the process is digging into complex information and uncovering the hidden patterns underneath.我擅长用 Python、R 及其他统计软件,把杂乱的数据变成有价值的决策依据。从数据清洗、建立模型、到挖掘有价值的信息,我享受着从海量复杂数据里发掘隐藏的规律。

I act as the bridge between what users want and what engineering can build. I regularly map out business workflows, design interactive prototypes in Figma, and write clear, structured PRDs. By keeping documentation organized and solutions practical, I help teams turn ideas into shipped features smoothly.我擅长在用户需求与技术可行性之间搭建沟通桥梁,能够独立撰写标准化 PRD、使用 Figma 梳理业务流程并输出交互原型,依托清晰规范的文档输出可落地的产品方案,推动需求高效落地上线。

I spent three years at my university's "Ji+" Media Center, eventually stepping up as Director and earning the Top 10 Visual Journalist award twice. Photography is how I tell stories and make sense of the world. Driven by an endless curiosity, I'm always looking for new ways to capture life through my own unique visual lens.我在学院"极+"融媒体中心任职3年,曾任主任一职,并连续两年获评十佳视觉记者。我热爱摄影,以镜头为载体讲述我眼中的世界。我始终保有旺盛好奇心,形成独属于自己的视觉创作视角。

Away from the screen, I love outdoor sports and fitness. For me, the dopamine release after a good sweat does more than just keep me physically fit—it recharges my brain. This energetic lifestyle keeps my mind sharp, allowing me to tackle work challenges with ease and truly enjoy every single moment of life.除了面对屏幕,我还喜爱户外运动与健身。对我来说,大汗淋漓后多巴胺的释放不仅能强健体魄,更为大脑充电续航。这种充满活力的生活方式让我始终保持思维敏捷,不仅能游刃有余地应对工作中的重重挑战,更能让我全心全意地享受生活的每一个瞬间。

02

Skills技能 Universe世界

WHERE ANALYTICAL THINKING MEETS CREATIVE EXPRESSION理性分析与创意表达在此交汇

01

Data Analytics数据分析

FIELD NOTES / SKILLS领域笔记 / 技能
Python
R Language
MATLAB
Excel
LSTM
XGBoost
NLP
Python
R Language
MATLAB
Excel
LSTM
XGBoost
NLP
COLLECTION合集
02

Product & Design产品与设计

FIELD NOTES / SKILLS领域笔记 / 技能
PRD Writing
UX Research
Data Storytelling
User Flows
Wireframing
Interaction Design
PRD Writing
UX Research
Data Storytelling
User Flows
Wireframing
Interaction Design
COLLECTION合集
03

Research Methods研究方法

FIELD NOTES / SKILLS领域笔记 / 技能
Monte Carlo Simulation
Structural Equation Modeling
Text Mining & NLP
Quantile Regression
Markov Chains
Field Research & Survey
Monte Carlo Simulation
Structural Equation Modeling
Text Mining & NLP
Quantile Regression
Markov Chains
Field Research & Survey
COLLECTION合集
04

Tools工具

FIELD NOTES / SKILLS领域笔记 / 技能
Python
R
VS Code
GitHub
Figma
PS
LR
EXCEL
Power BI
Power Query
Python
R
VS Code
GitHub
Figma
PS
LR
EXCEL
Power BI
Power Query
COLLECTION合集

English Proficiency英语能力

7.0 IELTS AcademicIELTS 学术
321 GRE

Photography & Media摄影与媒体

Passionate about photography热爱摄影
2x Top 10 Visual Journalist Award两次获评十佳视觉记者
Director of Converged Media Center融媒体中心主任

Photography Portfolio摄影作品集

Capturing moments, telling stories through my lens用镜头捕捉瞬间,讲述每一个故事

01

Dancing with the Stars: The Mirrorball's Destiny《与星共舞》数学建模

+

Investigating the conflict between professional elitism and mass public appeal in reality TV voting systems.探究真人秀投票机制中,专业评委与大众观众的娱乐偏好之间常常存在冲突。

Overview & Background概述与背景

The show "Dancing with the Stars" struggles with the conflict between expert technical scoring and subjective fan voting, often leading to controversial elimination results.《与星共舞》节目面临着专家技术打分与粉丝主观投票之间的冲突,因此往往会引发出具有争议的淘汰结果。

To develop a mathematical model that estimates hidden fan voting data, evaluates past scoring rules, and proposes a fairer voting mechanism.建立数学模型来估算隐藏的粉丝投票数据,来评估过去的评分规则,并提出一个更公平的投票机制。

As the core modeler on the team, I led the design and code implementation of the underlying algorithmic framework. I was responsible for building a vote estimation model using Monte Carlo algorithms and translating complex audience psychology into quantifiable, interpretable mathematical features through Random Forest and SHAP attribution analysis in machine learning.作为团队中的核心建模手,我主导底层算法框架的设计与代码实现。负责运用蒙特卡洛算法构建投票估算模型,并通过机器学习中的随机森林与SHAP归因分析,将复杂的受众心理转化为可量化、可解释的数学特征。

Core Methodologies核心方法论

  • PL-PECON: Used Monte Carlo rejection sampling to decode latent fan preferences, achieving a 96.73% historical consistency rate.使用蒙特卡洛拒绝采样解码潜在的粉丝偏好,并实现96.73%的历史一致性。
  • PRIME-AMC: Used Markov Chains to evaluate and contrast the scoring effects of ranking versus percentage methods.使用马尔可夫链来评估与对比排名制与百分比制的打分效果。
  • DSRF-SHAP: Combined SHAP with a Dual-Stream Random Forest model to decouple voting drivers.结合 SHAP 的双流随机森林模型,解耦投票驱动因素。
  • SARF: A Sigmoid-gated adaptive rank fusion that conditionally activates fan votes based on technical thresholds.基于技术阈值条件激活粉丝投票的Sigmoid门控自适应排名融合。

Research Findings研究结论

When a percentage-based voting method is adopted, there is a severe bias towards fan votes. Conversely, the ranking method effectively suppresses extreme variance and maintains a high signal-to-noise ratio of 0.96.当采取百分制投票时,粉丝投票倾向严重,而排名制能有效抑制极端方差,并保持 0.96 的高信噪比。

Recommended a hybrid strategy combining the ranking method with a "Judges' Save" mechanism, supplemented by dynamic gating, to ensure the trophy truly represents dancing excellence.建议采取混合策略,即排名制+评委保障机制,并辅以动态门控,以确保奖杯真正代表卓越的舞蹈水平。

02

Hainan Tourism Research: Predicting High-Engagement Guides海南旅游研究:高点赞旅游攻略预测

+

Revealing the core characteristics of highly-liked user travel guides on Qunar.com through data-driven analysis.基于数据驱动分析,揭示去哪儿网上高赞用户旅游攻略的核心特征。

Overview & Background概述与背景

Although highly-liked travel guides on the internet profoundly influence modern travel planning, it is difficult for both content creators and platforms to accurately identify what factors drive this high engagement.尽管网络上高点赞旅游攻略深刻影响着现代出行计划,但内容创作者和平台方都很难精准识别究竟是什么因素驱动了这些高互动量。

To build predictive models using machine learning and natural language processing techniques, identify key drivers of "likes," and provide actionable strategies.运用机器学习和自然语言处理技术构建预测模型,识别点赞的关键因素,并提供可落地的策略。

As the lead data analyst, I independently completed the entire data mining process, from unstructured text processing to feature engineering. I proficiently utilized NLP techniques to clean and parse massive amounts of text, and built Decision Tree and Quantile Regression models to transform discrete text content into predictive metrics.作为数据分析主力,我独立完成了从非结构化文本处理到特征工程的全流程数据挖掘。熟练运用 NLP 技术清洗并解析海量文本,并构建决策树与分位数回归模型,将离散的文本内容转化为预测指标。

Core Methodologies核心方法论

  • Semantic Analysis & NLP: Applied text mining and LDA topic modeling to extract structured features from unstructured travelogues.运用文本挖掘和LDA主题模型,从非结构化的游记攻略中提取结构化特征。
  • Quantile Regression: Analyzed the non-linear impact of per capita consumption, revealing a negative correlation in low-to-medium engagement content, but an insignificant impact in top-tier, highly-liked content.分析人均消费的非线性影响,揭示其在中低互动量内容中呈负相关,但在头部高赞内容中影响不显著。
  • Decision Tree Classification: Built and optimized a classifier using grid search to predict the likelihood of high engagement based on dimensions such as travel group type, attractions, and number of photos.构建并使用网格搜索优化分类器,基于出行群体类型、景点及照片数量等维度来预测高互动的可能性。

Research Findings研究结论

"Shopping" (0.550) and "Food" (0.160) are the strongest predictors, with weights far exceeding the mere mention of popular beaches. Furthermore, travelogues featuring "family" or "friends" traveling together have higher engagement rates."购物"(0.550)和"美食"(0.160)是最强的预测指标,其权重远超单纯提及热门海滩。此外,"家庭"或"朋友"结伴出行的游记互动率更高。

Formulated a content creation formula (e.g., combining in-depth tours, food, and photography for group travel) to maximize engagement rates and optimize the platform's recommendation algorithms.总结出内容创作公式(例如:针对团队游,将深度游、美食与摄影相结合),以最大化互动率并优化平台的推荐算法。

03

Zhejiang Entrepreneurship & Organizational Resilience浙商精神与组织韧性

+

Employing a mixed-methods research approach to deeply explore the impact of regional business culture on corporate green transition and organizational resilience.采用混合研究方法,深入探讨地域商业文化对企业绿色转型与组织韧性的影响。

Overview & Background概述与背景

Against the backdrop of the "Dual Carbon" goals, traditional manufacturing enterprises face severe challenges during their green transition.在"双碳"目标背景下,传统制造企业在绿色转型过程中面临严峻挑战。

To investigate how the "Zhejiang Entrepreneurship Spirit" drives corporate value creation and enhances organizational resilience using empirical data and case studies.利用实证数据和案例研究,探究"浙商精神"如何驱动企业的价值创造并提升组织韧性。

I was deeply involved in field research and data analysis. I performed multi-level coding on 60,000 words of in-depth interview transcripts, achieving a perfect integration of data models and business narratives.我深度参与了实地调研与数据分析,对6万字的深度访谈记录进行多级编码,实现了数据模型与商业叙事的完美结合。

Core Methodologies核心方法论

  • Empirical Analysis: Conducted a questionnaire survey of 264 startups. Used Structural Equation Modeling (SEM) to quantify impact paths, and applied Importance-Performance Analysis (IPA) to process interview texts.对共264家初创企业进行问卷调查。使用结构方程模型 (SEM) 量化影响路径,并运用重要性-表现分析法 (IPA) 处理访谈文本。
  • Longitudinal Case Study: Conducted an in-depth analysis of Hangang Group's 10-year transition journey, constructing the APCC (Antecedents-Process-Consequences-Chinese Context) model.深度剖析杭钢集团 10 年的转型历程,构建了APCC(前因-过程-结果-中国情境)模型。
  • Field Research: Conducted over 5 hours of in-depth interviews, compiled 60,000 words of verbatim transcripts, and used multi-level coding for qualitative comprehensive analysis.开展5小时以上的深度访谈,整理出6万字逐字稿,使用多级编码进行定性综合分析。

Research Findings研究结论

Validated that regional spirit can enhance corporate resilience by stimulating a willingness to take risks, promoting cooperation, alleviating financing constraints, and strengthening a sense of responsibility.验证了地域精神能够通过激发风险承担意愿、促进合作、缓解融资约束以及强化责任感来提升企业韧性。

Proposed a synergistic strategy combining an "efficient market" with a "proactive government" for policymakers, providing a roadmap for the transformation of the traditional manufacturing industry.为政策制定者提出了将"有效市场"与"有为政府"相结合的协同策略,为传统制造业转型提供路线图。

  • 0-to-1 Product Design产品设计

    Spearheaded the end-to-end requirement design for the core "Data Correction" module. Created high-fidelity Figma prototypes and comprehensive PRDs, streamlining the workflow by removing 30% of redundant features while introducing critical new capabilities like batch selection and draft saving.独立展开"数据纠错"模块的需求设计。通过Figma设计原型并撰写了详尽的 PRD 文档。通过剔除 30% 的冗余功能并引入批量选择、草稿保存等关键新功能,优化了工作流。

  • User-Centric Research以用户为中心的调研

    Conducted on-site field visits and 1-on-1 interviews with front-line users to identify usability blockers. Synthesized findings into 8 core business pain points, providing data-driven evidence that guided the optimization of multi-level approvals and batch operations.开展了实地走访,调研系统的使用体验;并收集对移动端场景的使用需求。提炼8大核心痛点,与UI对接PC 端交互改版设计。并形成移动端需求方案纳入产品迭代排期,跟进项目迭代升级。

  • Workflow Standardization工作流标准化

    Deconstructed and mapped a complex 5-level approval workflow involving 6 distinct user roles. Authored over 10 standardized operational manuals, successfully unifying system terminology and significantly reducing training costs for grassroots users.拆解并梳理了涉及 6 种不同用户角色的复杂 5 级审批工作流。撰写了 10 余份标准化操作手册,统一了系统术语,为基层填报减负。

  • Cross-Functional Agile Delivery团队合作与交付

    Actively participated in requirement and QA review meetings, aligning understanding across UI, development, and testing teams. Tracked project schedules and facilitated continuous version iterations to ensure seamless feature deployment.积极参与需求评审和测试用例评审会议,确保 UI、开发和测试团队的认知对齐。跟踪项目进度,推动版本的持续迭代,以保障功能的顺利上线。

  • Low-Level System Vulnerability Resolution底层系统漏洞攻坚

    Delved into memory management and network packet processing mechanisms. Assisted the development team in accurately locating and patching high-risk vulnerabilities at the system level, such as memory overflows and double-free errors, significantly enhancing the robustness and operational security of the system architecture.深入剖析内存管理与网络数据包处理机制,协助开发团队精准定位并修复了系统底层的内存溢出及重复释放等高风险漏洞,显著提升了系统架构的健壮性与运行安全性。

  • Network Threat Hunting & Traceability网络攻击威胁溯源

    Participated in the identification, investigation, and tracing of network traffic attack incidents. By extracting and analyzing abnormal traffic features, assisted in building an analysis pipeline for attack behaviors, delivering highly valuable business solutions for subsequent defense rule formulation and emergency response plans.参与网络流量攻击事件的识别、排查与溯源工作。通过提取和分析异常流量特征,协助构建了攻击行为的分析链路,为后续防御规则的制定与应急响应方案提供了极具业务价值的解决方案。

  • Macroeconomic Trend Monitoring宏观经济态势监测

    Participated in the daily dynamic monitoring and risk early-warning mechanisms for county-level economic operations. Integrating regional industrial characteristics with historical cycle data, assisted in producing economic performance analysis reports to provide objective and timely data references for macroeconomic regulation and policy formulation by relevant departments.参与执行县域经济运行态势的日常动态监测与风险预警机制。结合区域产业特征与历史周期数据,辅助输出经济运行分析材料,为相关部门的宏观调控与政策制定提供客观、及时的数据参考。

  • Multi-Source Government Data Governance多源政务数据治理

    Responsible for aligning and processing fundamental macroeconomic indicators. Efficiently completed the collection, cleaning, multi-dimensional cross-verification, and statistical analysis of complex data. Applied strict data validation logic to ensure the absolute accuracy and consistency of core government data.负责对接并处理来自宏观经济基础指标。高效完成各项复杂数据的收集、清洗、多维交叉核对与统计分析,通过严密的数据校验逻辑,确保了核心政务数据的绝对准确性与一致性。

I'm always excited to connect with like-minded individuals and explore new opportunities. Whether you have a project in mind, want to discuss data and product, or just want to say hi—feel free to reach out!我始终期待与志同道合的朋友相遇,共同探索新的机会。无论你有什么项目想法,或想聊聊数据与产品,又或者只是想说声"你好"——都欢迎随时联系我!

Mobile电话 (+86) 139-6839-9619
Location所在地 Hangzhou, China中国 杭州
University学校 Zhejiang Gongshang University浙江工商大学

Scan to Connect扫码联系

Add me on WeChat to stay in touch扫描二维码,添加我的微信保持联系

WeChat QR Code