Research Programs科研项目

Faculty-Mentored Research Catalog科研项目目录

Spanning Mathematics, Computer Science, Biology & Medicine, Architecture, and Economics — each project is guided 1-on-1 by mentors from MIT, Harvard, Stanford, CMU, and other leading universities.覆盖数学、计算机、生物医学、建筑、经济商科等多个学科，来自 MIT / Harvard / Stanford 等名校的导师 1v1 带队，全程真实参与。

13Programs科研项目
7Disciplines学科方向
1v1Mentorship专属导师
13 Programs13 个项目
MATHEMATICS数学MIT MentorMIT 导师12 Weeks12 周

Introduction to Algebraic Topology: From Polygons to Surfaces拓扑学研究入门

Starting from foundational topological concepts, explore manifolds, group theory, and the fundamental group, progressing to core algebraic topology tools such as the Seifert–van Kampen theorem, and culminating in a 5-page expository paper and presentation on a chosen topic.从拓扑基础概念出发，学习流形、群论与基本群，逐步过渡到 van Kampen 定理等代数拓扑核心工具，最终自选主题完成一篇约 5 页的综述型论文并展示。

~5-Page Expository Paper约 5 页综述型论文Final Presentation课题展示Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules · Advanced Track Available6 个模块 · 含进阶方向
  1. Sessions 1–2 | Topological Foundations & Surfaces: Open sets, continuous maps, homeomorphisms; examples of surfaces and proving sets are manifolds第 1–2 课｜拓扑基础与曲面：开集、连续函数、双射、等价等基本概念，认识曲面的例子；练习证明某些集合是流形
  2. Sessions 3–4 | Euler Characteristic, Quotient Spaces & Surface Classification: RP², S², T², and identifying surfaces from polygon gluings via orientation and Euler characteristic第 3–4 课｜欧拉示性数、粘合空间与曲面分类：RP²、S²、T² 等粘合空间，从多边形模型识别曲面（定向与欧拉数）
  3. Sessions 5–6 | Group Theory & Structure: Groups, free products, cosets, homomorphisms, quotient groups, and classification of finitely generated abelian groups第 5–6 课｜群论与群的结构：群、自由积、陪集、群同态、商群，有限生成阿贝尔群的分类
  4. Sessions 7–8 | The Fundamental Group: Homotopy of paths and computing fundamental groups of manifolds and gluing spaces第 7–8 课｜基本群及其计算：定义基本群与同伦，计算前面学过的流形与粘合空间的基本群
  5. Sessions 9–10 | Van Kampen Theorem & Applications: Degree of maps, retractions, fundamental group of wedges of circles, and induced homomorphisms第 9–10 课｜基本群的工具与 van Kampen 定理：映射度、收缩核、圆周楔和的基本群、诱导同态
  6. Sessions 11–12 | Independent Topic & Presentation: Deep dive into a chosen topic, short oral presentation, and a ~5-page expository paper第 11–12 课｜课题与展示：自选感兴趣的主题深入学习并做简短展示，完成一篇约 5 页的综述型论文（expository paper）

Advanced Directions (tailored by mentor): Preparation for entry-level mathematical research problems in algebraic topology and manifolds进阶方向（按学生基础由导师安排）：学完后可进入代数拓扑与流形方向的入门级数学研究问题

Level & Fit: High school, undergraduate, and prospective graduate students interested in pure mathematics适合：对纯数学感兴趣的高中生、本科生及申请硕博的学生Inquire →预约咨询 →
MATHEMATICS数学MIT MentorMIT 导师12 Weeks12 周

Geometric Analysis & Partial Differential Equations几何分析与偏微分方程研究

Begin with Laplace and heat equations alongside differential geometry of curves and surfaces (Gauss–Bonnet theorem and minimal surfaces), then closely read the classic Gage–Hamilton paper on curve shortening flow under mentor supervision, reconstructing omitted proofs or extending a result.从拉普拉斯方程、热方程与曲线曲面的微分几何入手，接触 Gauss–Bonnet 定理与极小曲面，再在导师督导下精读曲线缩短流的经典论文（Gage–Hamilton），把文献中省略的推导补全，或尝试将某个结论推广一步。

Research Paper / Report项目 paper / 科研报告Final Presentation结课展示Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules · Advanced Track Available6 个模块 · 含进阶方向
  1. Weeks 1–2 | Laplace Equation & Harmonic Functions: Maximum principles, fundamental solutions, derivative estimates, and Liouville’s theorem第 1–2 周｜拉普拉斯方程与调和函数：从最大值原理出发推导基本解，做导数估计，认识 Liouville 定理
  2. Weeks 2–3 | Heat Equation & Parabolic Methods: Parabolic maximum principles, heat kernel, and a priori estimates第 2–3 周｜热方程与抛物方法：把椭圆方程的手法迁移到热方程，学习抛物最大值原理、基本解与导数估计
  3. Weeks 4–7 | Differential Geometry of Curves & Surfaces: Curvature, torsion, Theorema Egregium, and the Gauss–Bonnet theorem第 4–7 周｜曲线与曲面的微分几何：曲率与挠率、空间曲线基本定理、Theorema Egregium、Gauss–Bonnet 定理
  4. Week 8 | Minimal Surfaces (Plateau Problem): First and second variation formulas, stability, and Bernstein’s problem第 8 周｜极小曲面（肥皂膜问题）：第一、第二变分公式，稳定性判断，Bernstein 问题
  5. Weeks 9–10 | Curve Shortening Flow & Classic Paper Reading: Section-by-section close reading of Gage–Hamilton with mentor guidance第 9–10 周｜曲线缩短流与经典论文精读：在导师督导下逐段读懂 Gage–Hamilton 论文，复原关键论证
  6. Weeks 11–12 | Research Paper & Presentation: Reconstruct omitted derivations or extend a theorem, concluding with a final presentation第 11–12 周｜项目 paper：补全文献中省略的推导，或尝试把某个结论推广一步，并完成结课展示

Advanced Directions (tailored by mentor): Long-term mathematics research track aiming for an arXiv Preprint or manuscript (6+ months, individually scoped with mentor)进阶方向（按学生基础由导师安排）：以形成 Preprint 或论文稿为目标的长期数学研究（6 个月起，导师个体定题）

Level & Fit: High school and undergraduate students with calculus foundations exploring theoretical math research, as well as graduate applicants适合：有微积分基础、第一次接触理论数学研究的高中生与本科生，及申请硕博的学生Inquire →预约咨询 →
ARCHITECTURE建筑Harvard MentorHarvard 导师12 Weeks12 周

Architectural Design Research & Portfolio建筑设计研究与作品集

Trace modern Japanese architecture and Metabolism to investigate an architect’s critical position; perform digital surveying and scaled redrawing of a canonical house; then alter a single dimensional parameter to trace the resulting spatial, structural, and tectonic transformations.沿日本现代建筑与 Metabolism 的阅读线，研究一位建筑师的作者立场；对一栋经典住宅做数字测绘与重绘，标注真实尺寸比例；再只改变一个尺寸，追踪形式、空间、结构与建造随之发生的全部变化。

4-Board Final Review (Analysis + Redesign)四页最终展示（分析＋改造＋presentation）Portfolio Design Narrative作品集叙事（portfolio narrative）Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules6 个模块
  1. Modern Japanese Architecture Reading Seminar: Analyze three foundational texts on selective appropriation, distance, and modernization, writing critical responses日本现代建筑阅读线：读三篇导读，追踪人物、作品与「选择性挪用、距离、现代化」等概念，写出自己的评论
  2. Metabolism & Urban Systems: Situate architecture within frameworks of metabolic growth, urbanism, and technologyMetabolism 与城市系统：把建筑放进系统、增长、城市与技术的框架里讨论
  3. Architect’s Critical Position: Investigate the core disciplinary questions a selected architect responds to建筑师的作者立场研究：判断一位建筑师在回应什么问题
  4. Digital Surveying & Redrawing of a Canonical House: Produce accurately scaled drawings and renderings to test spatial and proportional understanding经典住宅的数字测绘与重绘：标注真实尺寸比例并完成渲染，检验对空间与尺度的理解
  5. Single-Variable Design Transformation: Modify one critical dimension and trace all resulting formal, spatial, structural, and constructional shifts单变量设计推演：只改变一个尺寸，追踪形式、空间、结构与建造随之发生的全部变化
  6. Final Review & Portfolio Integration: Complete a 4-page visual argument and translate it into a portfolio narrative with post-program mentor feedback最终展示：完成四页展示（分析＋改造＋presentation）并转化为作品集叙事；项目结束后导师继续给最终反馈，包括 PDF 修订与作品集调整
Level & Fit: Students applying to architecture and design programs seeking deeper critical rigor and narrative clarity in their portfolios适合：申请建筑及相关专业、希望为作品集补充思想深度与表达逻辑的学生Inquire →预约咨询 →
BIOLOGY & MEDICINE生物医学MIT MentorMIT 导师12 Weeks12 周

Cancer Immunotherapy & Vaccines Research肿瘤免疫治疗前沿研究

Starting from how the immune system recognizes tumors, systematically study immune checkpoint blockade, CAR-T / CAR-NK cell therapies, and cancer vaccines, then engage with primary literature on combination therapies to complete an independently scoped project.从"免疫系统如何识别肿瘤"出发，系统学习免疫检查点抑制、CAR-T / CAR-NK 细胞治疗与癌症疫苗，再进入联合疗法与前沿文献，自主定题完成研究。

Research Paper or PresentationResearch paper 或 presentation（二选一）Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排5 Modules · Advanced Track Available5 个模块 · 含进阶方向
  1. Foundations of Tumor Immunology: How innate and adaptive immunity recognize and eliminate malignant cells肿瘤免疫基础：建立「免疫系统如何识别肿瘤」的框架
  2. Immune Checkpoint Blockade: Mechanisms of PD-1/PD-L1 and CTLA-4 pathways that transformed clinical oncology免疫检查点抑制：理解这类已改变临床实践的疗法的机制
  3. Cellular Immunotherapy: Engineering principles and clinical boundaries of CAR-T and CAR-NK therapies细胞治疗：比较 CAR-T 与 CAR-NK 的构造思路与适用边界
  4. Cancer Vaccines: Neoantigen selection and vaccine delivery platforms from mechanism to therapeutic strategy癌症疫苗：理解抗原选择与疫苗平台如何从机制走向策略
  5. Combination Therapies & Independent Inquiry: Close reading of frontier papers, independent topic formulation, mini-presentation, and review paper联合疗法、前沿文献与自主选题：读前沿文献、自主定题，完成 mini presentation 与 review

Advanced Directions (tailored by mentor): mRNA Vaccine Design · Neoantigen Discovery · Protein Engineering进阶方向（按学生基础由导师安排）：mRNA 疫苗、抗原设计、蛋白工程

Level & Fit: Students interested in biomedicine, immunology, bioengineering, or translational medicine with foundational biology/chemistry knowledge适合：生物医学、免疫学、生物工程、转化医学方向，具备英文论文阅读能力和基础生物化学知识的学生Inquire →预约咨询 →
BIOLOGY & MEDICINE生物医学MIT MentorMIT 导师12 Weeks12 周

Stem Cell Biology & Disease Modeling干细胞生物学研究

Explore the differentiation mechanisms of human pluripotent stem cells (hPSCs), patient-specific disease modeling, and regenerative medicine applications through primary literature and clinical trials, culminating in a conference-style research proposal and presentation.围绕人多能干细胞的分化机制、疾病建模与再生医学应用展开，结合真实文献与临床试验案例，以国际学术会议的形式完成研究报告与展示。

4–6 Page Research Proposal4–6 页研究报告（研究计划）Conference-Style Presentation会议式课题展示Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排9 Modules9 个模块
  1. Session 1 | Introduction to Human Pluripotent Stem Cells: Properties and classification of embryonic, adult, and induced pluripotent stem cells (iPSCs)第 1 课｜人多能干细胞导论：干细胞的定义与特性，胚胎、成体与诱导多能干细胞的分类，干细胞在胚胎发育中的作用
  2. Sessions 2–3 | History & Derivation of hPSCs: Reprogramming milestones, somatic cell nuclear transfer, and laboratory maintenance第 2–3 课｜hPSC 的历史与获得：关键里程碑、重编程技术、实验室培养与维持；完成小测验
  3. Sessions 4–5 | Differentiation Potential & Organoids: Directed differentiation protocols and advanced tissue models; literature synthesis on an under-studied cell lineage第 4–5 课｜干细胞的分化潜能：定向分化方案与先进组织模型；为某一细胞类型的分化文献写摘要综述，找出一个研究较少的细胞类型并提出研究问题
  4. Sessions 6–7 | Disease Modeling & Therapeutics: Building patient-derived models and analyzing active stem cell clinical trials第 6–7 课｜干细胞与疾病建模：疾病分子机制、患者疾病模型的构建、作为疗法开发平台；总结一篇近期疾病建模论文和一项正在进行的干细胞疗法临床试验
  5. Session 8 | Bioethics & Regulatory Frameworks: Ethical considerations of hESCs, gene editing, and a 500-word perspective piece第 8 课｜伦理与法律：胚胎干细胞的伦理问题、监管框架、基因编辑等新技术；写一篇 500 词观点文章
  6. Session 9 | Research Journal Club: Critical discussion of an impactful recent stem cell publication第 9 课｜研究 Journal Club：研读并讨论近期有影响力的干细胞研究论文
  7. Session 10 | Bench-to-Bedside Translation: Identifying unmet clinical needs addressable via stem cell platforms第 10 课｜回顾与展望：未来应用、从研究走向临床的挑战与机会；找出可由干细胞研究解决的未满足临床需求
  8. Session 11 | Research Topic Formulation: 1-on-1 scoping of hypothesis, specific aims, and experimental design第 11 课｜确定研究课题：与导师讨论并确定每位学生的最终课题与产出
  9. Session 12 | Final Symposium Presentation: Deliver a 4–6 page research proposal and oral presentation in international conference format第 12 课｜结项展示：撰写研究计划（研究问题的依据、研究目标与实验设计、预期结果），完成 4–6 页报告，以国际学术会议的形式展示
Level & Fit: High school, undergraduate, and prospective graduate students interested in cell biology, regenerative medicine, and biomedicine适合：对生物医学、细胞生物学感兴趣的高中生、本科生及申请硕博的学生Inquire →预约咨询 →
BIOLOGY & MEDICINE生物医学Monash MentorMonash 导师12 Weeks12 周

Medical Imaging Principles & AI Analysis医学影像原理与 AI 分析

Compare X-ray, CT, MRI, and ultrasound imaging physics, understand signal acquisition and tomographic reconstruction, implement image filtering, segmentation, and feature extraction in Python (scikit-image & SimpleITK), and formulate an independent AI imaging study.比较 X-ray、CT、MRI、超声四大成像模态，读懂磁场如何变成信号、投影如何重建成图，读一篇 PET 或 fMRI 近期论文，用 scikit-image 与 SimpleITK 做图像滤波、分割与特征提取，最后自定研究问题。全程在电脑上完成。

4–6 Page Research Report4–6 页研究报告Final Research Presentation研究展示
View Syllabus & Schedule查看项目安排9 Modules · Advanced Track Available9 个模块 · 含进阶方向
  1. Session 1 | Modalities Overview: Physics and clinical trade-offs across X-ray, CT, MRI, and Ultrasound第 1 课｜医学影像导论：X-ray、CT、MRI、超声的原理、临床用途与基础物理；写一份各模态优势与局限的比较总结
  2. Sessions 2–3 | MRI & CT Reconstruction: Magnetic resonance signal processing, k-space, and CT back-projection in neuro and oncology imaging第 2–3 课｜MRI 与 CT 原理：磁场与信号处理、CT 图像重建，神经与肿瘤领域的临床应用；完成测验
  3. Sessions 4–5 | Ultrasound & Radiography: Acoustic physics, 2D vs. 3D ultrasound, and Doppler flow analysis第 4–5 课｜超声与 X-ray：超声成像物理、X-ray 在骨科与肺部疾病中的应用；比较 2D 与 3D 超声，分析多普勒超声的应用
  4. Sessions 6–7 | Functional & Molecular Imaging: PET, SPECT, fMRI, multimodal fusion, and close reading of a recent paper第 6–7 课｜高级成像技术：PET、SPECT、fMRI 与图像融合；精读一篇近期 PET 或 fMRI 论文并总结
  5. Session 8 | Computational Image Analysis: Filtering, segmentation, and feature extraction in Python using scikit-image and SimpleITK第 8 课｜图像处理与分析：滤波、分割、特征提取，机器学习在医学影像中的应用；用 scikit-image 或 SimpleITK 完成一个图像处理任务
  6. Session 9 | Clinical AI Ethics & Regulation: Patient data privacy, FDA/clinical validation, and a 500-word ethics brief第 9 课｜伦理与法律：数据隐私、临床使用的监管规范；写一篇约 500 词的 AI 医学影像伦理短文
  7. Session 10 | AI Medical Imaging Journal Club: Critical evaluation of a deep-learning radiology publication第 10 课｜研究 Journal Club：研读一篇 AI 医学影像论文并讨论
  8. Session 11 | Research Scoping: Define a computational imaging research question and experimental pipeline第 11 课｜确定研究课题：提出研究问题，讨论可能的实验与方法
  9. Session 12 | Final Report & Defense: Present a 4–6 page written report and research presentation第 12 课｜结项展示：完成 4–6 页报告并展示拟议研究

Advanced Directions (tailored by mentor): Accelerated MRI Acquisition & Reconstruction: Parallel imaging, compressed sensing, low-rank models, and physics-informed deep learning进阶方向（按学生基础由导师安排）：MRI 快速采集与图像重建：并行成像、压缩感知、低秩重建、物理约束的深度学习重建

Level & Fit: Students interested in biomedical engineering, medical imaging, electrical engineering, computer vision, and AI in healthcare适合：生物医学工程、医学影像、电子信息、计算机与 AI 方向，以及想学医但更想做技术的学生Inquire →预约咨询 →
ECON & BUSINESS经济商科NYU MentorNYU 导师12 Weeks12 周

Economic Growth & Macroeconomic Data Analysis经济增长与宏观数据研究

Compare cross-country income using Purchasing Power Parity (PPP), uncover empirical regularities from real macroeconomic datasets, calibrate the Solow growth model to selected countries, and incorporate human capital, innovation, and inequality metrics into a data-driven research paper.用购买力平价比较各国收入，从真实国家数据中整理经验规律，把 Solow 增长模型套到自选国家上做预测，再加入人力资本、创新与不平等指标，完成一份有数据、有模型、有图表的研究。

5-Page Empirical Research Report5 页研究报告20-Minute Research Presentation约 20 分钟研究展示Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules6 个模块
  1. Sessions 1–2 | Cross-Country Comparability: Nominal vs. real GDP, Purchasing Power Parity (PPP), and assembling national macroeconomic datasets第 1–2 课｜指标可比性：名义量与实际量、购买力平价（PPP），下载自选国家的宏观变量
  2. Sessions 3–4 | Empirical Stylized Facts: Data cleaning, visualization, and structural comparisons across income tiers第 3–4 课｜从数据中找经验规律：清理、筛选、整理国家数据，比较富国与较贫穷国家
  3. Sessions 5–6 | The Solow Growth Model: Capital accumulation, steady-state dynamics, and country-specific growth forecasting第 5–6 课｜Solow 增长模型：资本积累与稳态，用模型对自选国家做增长预测
  4. Sessions 7–8 | Human Capital, R&D & Endogenous Growth: Incorporating education, innovation, and total factor productivity into empirical forecasts第 7–8 课｜人力资本、创新与知识产权：加入人力资本与技术进步指标，更新增长预测
  5. Sessions 9–10 | Growth, Poverty & Inequality: Analyzing Gini and distributional metrics while distinguishing correlation from causality第 9–10 课｜增长与不平等：收集贫困与不平等指标，分析它们与增长的关系，区分相关与因果
  6. Sessions 11–12 | Empirical Synthesis: Complete a 5-page quantitative report and a 20-minute academic presentation第 11–12 课｜研究表达：汇总数据来源、理论依据与实证分析，完成 5 页报告与约 20 分钟展示
Level & Fit: Students interested in economics, development studies, public policy, finance, and quantitative data analysis适合：经济学、发展研究、公共政策、金融与数据分析方向的学生Inquire →预约咨询 →
ECON & BUSINESS经济商科USC MentorUSC 导师12 Weeks12 周

Global Marketing Communication Strategy全球营销传播研究

Centered on the tension between global standardization and local adaptation, examine market selection, entry modes, pricing, distribution, global brand architecture, and cross-cultural consumer behavior to produce a comprehensive case-based strategy report.围绕"全球一致性与本地适配"这条主线，研究市场选择与进入方式、价格与渠道、全球品牌的建立，以及文化如何改变广告诉求与消费者行为，完成一个全球品牌或市场进入方向的项目。

Final Case Research ReportFinal Project ReportFinal Project PresentationFinal Project PresentationFaculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules6 个模块
  1. Weeks 1–2 | The Global Marketing Environment: Trade, socio-cultural dynamics, and standardization vs. local adaptation第 1–2 周｜全球化环境：全球经济、贸易、社会与文化环境，以及「全球一致性与本地适配」的张力
  2. Weeks 3 & 6 | Market Selection, Entry Modes & Research: Segmentation, positioning, licensing, joint ventures, and foreign direct investment第 3、6 周｜市场选择、进入与市场研究：细分、目标市场与定位，比较进出口、许可、投资与战略联盟等进入方式
  3. Weeks 4–5 | Global Pricing, Distribution & Brand Equity: Cross-border channel architecture and building a global brand第 4–5 周｜价格、渠道与全球品牌：定价与分销路径，拆解一个全球品牌如何建立
  4. Week 7 | Competitive Advantage & Global Strategy: Evaluating firm-level strategic positioning across regions第 7 周｜竞争优势与全球战略：判断一家企业的竞争优势由什么构成
  5. Weeks 8–10 | Culture, Integrated Marketing Communications & Consumer Behavior: How cultural dimensions shape advertising appeals and media strategy第 8–10 周｜文化、广告与消费者行为：文化维度如何改变广告诉求、媒体选择、公关与促销
  6. Weeks 11–12 | Capstone Project: Final written research report and formal presentation第 11–12 周｜最终项目：Final Project Report 与 Final Project Presentation
Level & Fit: Students interested in communications, marketing, brand strategy, consumer behavior, and international business适合：传播学、营销、品牌、公关、消费者行为与全球商业方向的学生Inquire →预约咨询 →
COMPUTER SCIENCE计算机Stanford MentorStanford 导师12 Weeks12 周

AI for the Subsurface & Earth Sciences人工智能与地球科学应用研究

Build deep learning models in Python (Google Colab) alongside foundational geoscience concepts, training neural networks on real well-log data for lithology classification or applying CNNs to seismic sections for fault interpretation, concluding with a formal scientific report.在 Google Colab 里学习 Python 与深度学习框架，补上板块构造、岩石类型等地学常识，再用真实测井数据训练神经网络判断岩性，或用 CNN 从地震剖面识别断层，最后写成规范的英文科学报告。

Scientific Report with Figures岩性分类或断层解释的 scientific reportReproducible Code & Model Outputs可演示的代码与结果图表Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排6 Modules · Advanced Track Available6 个模块 · 含进阶方向
  1. Sessions 1–2 | Deep Learning Architectures: Neural networks, RNNs, CNNs, classification, regression, and generative modeling第 1–2 课｜深度学习概念与应用：神经网络、RNN、CNN，回归、分类与生成，拆解一个典型 AI 应用
  2. Sessions 3–4 | Python & Deep Learning Workflows: Building, training, and evaluating neural networks in Google Colab / Jupyter第 3–4 课｜Python 与深度学习框架：在 Google Colab / Jupyter 中搭建并训练小型网络，画图、计算准确率
  3. Sessions 5–6 | Geoscience Foundations: Plate tectonics, rock formations, stratigraphy, and baseline lithology classification code第 5–6 课｜地球科学背景：板块构造、岩石类型、地质年代，跑通一份岩性分类代码
  4. Sessions 7–8 | Well-Log Lithology Classification: Training neural networks on real borehole sensor curves to infer subsurface rock types第 7–8 课｜真实测井数据岩性分类：用测井曲线训练神经网络，判断地层岩性
  5. Sessions 9–10 | Seismic Fault Interpretation with CNNs: Applying convolutional networks to 2D/3D seismic profiles to detect geological faults第 9–10 课｜地震数据断层解释：用 CNN 从地震剖面识别断层与地层界面（部分代码由导师提供）
  6. Sessions 11–12 | Scientific Writing & Peer-Review Standards: Structuring abstracts, methodology, figures, and discussion in journal format第 11–12 课｜科学报告与英文写作：报告结构、摘要、引言、方法，并了解期刊审稿流程

Advanced Directions (tailored by mentor): Generative Geological Modeling · Physics-Informed Neural Networks (PINNs) · Surrogate Modeling & Uncertainty Quantification for Subsurface Flow (CO₂ storage, geothermal energy, groundwater)进阶方向（按学生基础由导师安排）：生成式地质建模、物理信息神经网络、地下流动的代理模型与不确定性分析（应用于二氧化碳地质封存、地热、地下水与天然氢）

Level & Fit: Students interested in AI for Science, computational geoscience, energy, and environmental engineering (no prior geology background required)适合：想进入 AI for Science、地球科学、能源与环境工程方向的学生，无需预先具备系统地学或高阶数学训练Inquire →预约咨询 →
COMPUTER SCIENCE计算机CMU MentorCMU 导师12 Weeks12 周

Machine Learning & AI: Models, Games, and Privacy人工智能与机器学习研究

Progress from supervised and unsupervised learning, gradient descent, neural networks, SVMs, and K-means clustering to combinatorial game-playing algorithms and privacy-preserving machine learning, guided by a mentor specializing in cryptography, security, and privacy.从监督与无监督学习、回归与分类讲起，经过梯度下降、逻辑回归、神经网络、SVM 与 K-means 聚类，再进入组合博弈的自动对弈算法和隐私保护机器学习。导师长期从事密码学、安全与隐私研究，可按学生基础把隐私保护方向讲深。

Machine Learning Project Report机器学习课题研究报告Algorithmic Implementation算法实践成果Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排5 Modules · Advanced Track Available5 个模块 · 含进阶方向
  1. Foundations of AI & ML: Supervised vs. unsupervised learning, classification, and regression formulations导论：什么是机器学习与人工智能、现实应用、监督与无监督学习、分类与回归
  2. Linear & Polynomial Regression: Loss functions, gradient descent optimization, and multivariate matrix methods线性与多项式回归：代价函数、梯度下降、多特征与矩阵方法，同步复习必要的线性代数
  3. Core Machine Learning Models: Logistic regression, neural network fundamentals, Support Vector Machines (SVM), and K-means clustering常见模型：逻辑回归与 sigmoid 函数、神经网络基础、支持向量机（SVM）、K-means 聚类
  4. Automated Game Playing: Combinatorial game theory, minimax search, and algorithmic strategies in board games自动博弈：组合博弈的基本理论与可实现的算法，以棋类游戏为例
  5. Privacy-Preserving Machine Learning: Data privacy risks and cryptographic techniques for secure model training and inference隐私保护机器学习：数据隐私问题、隐私保护机器学习的基本思路及其中用到的密码学技术

Advanced Directions (tailored by mentor): Cryptographic Protocols for Privacy-Preserving ML · Formal Definitions of Security & Differential Privacy · Primary Research Literature Reading进阶方向（按学生基础由导师安排）：隐私保护机器学习中的密码学方法、安全与隐私的形式化定义、相关研究论文精读

Level & Fit: High school, undergraduate, and prospective graduate students interested in ML theory, algorithms, and data privacy (necessary linear algebra reviewed in class)适合：具备基础数学、对 AI 原理、模型与数据隐私感兴趣的高中生和本科生，及申请硕博的学生（所需数学会在课上简要复习）Inquire →预约咨询 →
COMPUTER SCIENCE计算机CMU MentorCMU 导师12 Weeks · 12 Hours12 周 · 12 小时

Modern Cryptography Theory & Security Proofs密码学理论与研究报告

Start by breaking classical ciphers, build the number-theoretic foundations (modular arithmetic, discrete logarithms, DDH), analyze RSA, Diffie–Hellman, and digital signatures, and explore TLS/HTTPS, blockchain consensus, and zero-knowledge proofs to complete an original research report.从 Caesar、Vigenère 等古典密码出发，补齐模运算、离散对数等数论工具，理解 RSA、数字签名、Diffie–Hellman 为什么安全，再接触 HTTPS 背后的协议、区块链机制与零知识证明，完成一份研究报告。导师长期从事现代密码学与理论计算机研究，基础扎实的学生可以进一步走进安全证明与量子密码学。

Cryptography Research Report密码学研究报告Focused Study (Blockchain / Digital Signatures / Historical Cryptanalysis)课题方向三选一（区块链机制 / 数字签名 / 二战古典密码）Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排7 Modules · Advanced Track Available7 个模块 · 含进阶方向
  1. Classical Ciphers & Perfect Secrecy: Caesar, Vigenère, substitution ciphers, the one-time pad, and cryptanalytic attacks古典与现代密码：固定移位、Caesar、Vigenère、替换密码、one-time pad，以及它们的安全性分析
  2. Number-Theoretic Foundations: Modular arithmetic, prime factorization, discrete logarithms, and the Decisional Diffie–Hellman (DDH) assumption数学基础：模运算、数论、离散对数、因子分解、DDH 假设
  3. Modern Cryptographic Primitives: Collision-resistant hashing, MACs, digital signatures, Diffie–Hellman key exchange, ElGamal, and RSA现代密码原语：哈希、数字签名、认证、对称加密、Diffie–Hellman、公钥加密、ElGamal、RSA
  4. Internet Security Protocols: Public-key infrastructure (PKI), digital certificates, TLS/SSL, and HTTPS互联网安全：数字证书、SSL 与 HTTPS
  5. Blockchain & Distributed Consensus: Bitcoin architecture, Proof-of-Work vs. Proof-of-Stake, and open research challenges区块链：比特币、挖矿、共识机制、工作量证明与权益证明，加密货币的局限与活跃研究问题
  6. Advanced Topics: Zero-Knowledge Proofs (ZKPs) and cryptographic coin-flipping protocols高阶主题：零知识证明、线上 coin-flipping
  7. Research Project: Complete a formal research report on blockchain mechanisms, signature efficiency/security, or WWII Enigma cryptanalysis研究项目：从导师建议的方向中选题（区块链机制、数字签名的安全性与效率、二战古典密码破译），完成研究报告

Advanced Directions (tailored by mentor): Formal Security Definitions & Reduction Proofs · Adversarial Threat Modeling · Introduction to Post-Quantum & Quantum Cryptography · Reading Seminar on Crypto Papers进阶方向（按学生基础由导师安排）：形式化安全定义与安全证明、攻击模型与安全模型的建立、量子密码学入门、密码学研究论文精读

Level & Fit: High school, undergraduate, and graduate-track students interested in theoretical computer science, discrete mathematics, and mathematical proofs适合：有基础数学背景、对密码学与理论计算机感兴趣、愿意接受证明性内容的高中生与本科生；也适合数学、算法基础扎实、准备走理论研究路线的本科高年级与硕博申请者Inquire →预约咨询 →
ENVIRONMENTAL SCI环境科学Stanford MentorStanford 导师12 Weeks12 周

Environmental Nanoscience & Nanomaterials环境科学与纳米材料研究

Investigate how natural and engineered nanoparticles form, transform, and interact with contaminants across water, soil, and air; master advanced spectroscopy and microscopy characterization methods; and propose a nanomaterial-based solution to a real environmental or public health challenge.学习纳米颗粒在水、土壤、大气中的行为及其环境与健康影响，掌握散射、光谱等表征技术，独立或分组完成一份纳米材料应用于真实环境或健康问题的研究报告与展示。

Research Report & Proposal科研报告/课题展示Nanomaterial Application Analysis纳米材料应用分析Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排7 Modules7 个模块
  1. Sessions 1–2 | Nanoparticle Formation & Transformation: Natural vs. anthropogenic nanomaterials and their environmental fate第 1–2 课｜纳米颗粒的形成与转化：天然与人为纳米颗粒如何形成，在不同环境中如何转化
  2. Sessions 2–4 | Nanoclays & Iron Oxyhydroxides: Crystal structure, surface reactivity, and roles in soil and aquatic geochemistry第 2–4 课｜纳米黏土与铁羟基氧化物纳米颗粒：性质、结构及其在土壤与水化学中的作用
  3. Sessions 5–6 | Aqueous Contaminant Interactions: Adsorption, dissolution, and precipitation mechanisms for water remediation第 5–6 课｜纳米颗粒与水中污染物：吸附、溶解与沉淀机制，对水处理与污染控制的意义
  4. Sessions 7–8 | Atmospheric Aerosols & Air Quality: Sources, transport, and toxicity of airborne ultrafine particles第 7–8 课｜纳米颗粒对空气质量的影响：大气中纳米颗粒的来源、行为与毒性
  5. Sessions 9–10 | Advanced Characterization: Synchrotron scattering, spectroscopy, and electron microscopy (TEM/SEM)第 9–10 课｜先进表征技术：实验室与同步辐射的散射、光谱与成像方法（含电子显微镜）
  6. Sessions 10–12 | Nanotoxicology & Human Health: Exposure pathways, pulmonary impacts, and risk assessment第 10–12 课｜纳米颗粒的健康影响：毒性、暴露途径，对呼吸系统等的影响
  7. Capstone Project | Formulate a research report and intervention proposal addressing an environmental or health problem using a selected nanomaterial课题项目｜选择特定纳米颗粒，分析它在真实环境或健康问题中的作用并提出解决方案，涵盖科学背景、方案设计或预期结果、不确定性与风险三部分
Level & Fit: High school, undergraduate, and prospective graduate students interested in environmental science, chemistry, and materials engineering适合：对环境科学、材料科学感兴趣的高中生、本科生及申请硕博的学生Inquire →预约咨询 →
MEDIA & FILM传媒影视USC MentorUSC 导师12 Weeks12 周

Screenwriting & Short Film Directing Workshop剧本创作与短片制作研究

An intensive directing and screenwriting workshop deconstructing dramatic scene construction, narrative arc, visual storytelling, blocking, and camera coverage, guiding each student from an initial pitch through three rigorous drafts of an original short-film screenplay.精品小班导演与编剧工作坊，从场景设计、故事结构到视觉呈现层层拆解短片创作方法，最终独立完成一部短片剧本的反复打磨定稿。

Third-Draft Original Short Screenplay原创短片剧本（三轮修改）Directing & Visual Storytelling Craft导演/编剧方法论训练Faculty Recommendation (Merit-Based)导师推荐信（表现优秀）
View Syllabus & Schedule查看项目安排10 Modules10 个模块
  1. Session 1 | Function of a Scene I: Revealing character through action and subtext第 1 课｜场景的作用 I：塑造人物
  2. Session 2 | Function of a Scene II: Propelling narrative momentum第 2 课｜场景的作用 II：推动故事发展
  3. Session 3 | Story Architecture: Three-act structure, stakes, and dramatic conflict第 3 课｜故事基础：结构与冲突
  4. Session 4 | Visual Storytelling: Montage, dream sequences, and cinematic tone第 4 课｜讲故事的视觉：蒙太奇、梦境与风格
  5. Session 5 | Blocking & Coverage: Staging actors and planning camera setups for a scene第 5 课｜走位与镜头覆盖：如何调度演员、拍摄一场戏
  6. Session 6 | Frame Composition: Designing expressive shots and visual motifs第 6 课｜构建画面：如何设计一个好的镜头画面
  7. Session 7 | Deconstructing Short Films: Analyzing award-winning shorts and developing story concepts第 7 课｜拆解短片：观摩往届学生作品与优秀短片，构思提案
  8. Session 8 | The Pitch: Presenting loglines, character arcs, and dramatic premise第 8 课｜Pitch：提出短片剧本创意
  9. Session 9 | Story Workshop & First Draft: Table read and structural critique of Draft 1第 9 课｜创意工作坊：研讨创意，交出剧本第一稿
  10. Sessions 10–12 | Screenplay Workshops I–III: Page-by-page mentor notes through Draft 2 and polished Draft 3第 10–12 课｜剧本工作坊 I–III：逐稿研读、导师给出修改意见，依次完成第二、第三稿
Level & Fit: High school, undergraduate, and prospective graduate students interested in film production, screenwriting, and visual arts适合：对影视创作、编剧方向感兴趣的高中生、本科生及申请硕博的学生Inquire →预约咨询 →

How Mentoring Works标准流程

The 5-Step Research Experience5 步科研体验流程

  1. 01

    Academic Evaluation & Matching学术评估与课题匹配

    We evaluate your coursework, interests, and academic goals to match you with a suitable faculty mentor and research topic.根据学员背景、兴趣与目标，精准匹配名校导师与真实课题方向。

  2. 02

    Foundations & Methodology基础知识与研究方法学习

    Build essential domain foundations and learn the analytical, mathematical, or computational tools required for genuine inquiry.从学科基础概念讲起，逐步引入真实研究会用到的方法与工具，打好独立探索的底子。

  3. 03

    1-on-1 Guided Research1v1 科研实践与探索

    Weekly 1-on-1 online sessions where you conduct authentic research tasks — derivations, coding, data analysis, or design critique.每周 1v1 线上研讨，在导师带领下亲自动手做真实的研究任务——推导、编程、实验或案例分析，体验科研的真实节奏。

  4. 04

    Research Report Writing课题总结与报告撰写

    Write a structured academic research report or paper under your mentor’s guidance, synthesizing your methods and findings.在导师指导下独立完成一份规范的课题研究报告，把整个探索过程系统梳理成可展示的成果。

  5. 05

    Final Presentation & Outcome课题展示与学术履历转化

    Deliver a final presentation or defense. Strong performers may earn a faculty recommendation letter and a verifiable scholarly record.以展示或答辩形式呈现研究成果，表现优秀可获导师推荐信，成为申请材料中真实、可讲述的科研经历。

Academic Advisory学术评估与规划

Not Sure Which Program Fits Your Goals?不确定哪个方向适合你？

Speak with an academic advisor for a complimentary 1-on-1 background evaluation and personalized topic recommendation based on your grade, coursework, and target universities.联系学术顾问，免费获取 1v1 学术背景评估与课题匹配建议。告诉我们你的年级、专业与目标，我们为你量身定制科研路线图。

Schedule Free 1-on-1 Evaluation免费预约 1v1 学术评估
Begin Your Research Journey开启科研探索

Ready to Redefine Your Academic Profile Through Authentic Research?准备好用真实科研改写你的学术履历了吗？

Work one-on-one alongside university researchers to build verifiable scholarly outcomes in research and mathematics competitions.与全球名校导师同行，在科研与竞赛中建立不可替代的竞争壁垒。

Explore Research Programs探索核心服务方案Schedule 1-on-1 Evaluation预约 1v1 评估