I am Meng Zhichao, an algorithm engineer.
Starting from mathematics, I entered the data science field through self-study, progressing through power big data, financial risk modeling, NLP, and LLM engineering — gradually forming a working style centered on "model-driven thinking". I have published 200+ technical articles on CSDN with nearly 300,000 cumulative views.
In 2026, I participated in SemEval-2026 Task 7 (Multilingual Everyday Commonsense Evaluation) with a colleague. In Track 1 (Short Answer task), we ranked 1st with a score of 78.7672, and in Track 2 (Multiple Choice task), we achieved an overall accuracy of 96.35% on approximately 47,000 samples, ranking 2nd.
This repository contains notes from my daily learning and engineering practice, covering the following areas:
.
├── markdown_notes/ # Technical notes in Markdown format (approx. 2020 onwards)
├── handwritten_notes/ # Handwritten notes (scanned, approx. before 2020)
└── README.md
-
The
handwritten_notesfolder contains notes primarily from after university graduation up to around 2020, when handwriting was the main format. The content is math-heavy, covering calculus, linear algebra, statistical learning methods, machine learning, and foundational theory. Probability theory and mathematical statistics were also included, though those notes are no longer available. -
The
markdown_notesfolder contains notes primarily in Markdown format (converted to PDF inside the folder), covering learning and accumulated knowledge from work experience after 2020.
- All notes were written by me personally. The handwritten PDF scans have original notebooks as source, and the Markdown notes have original
.mdfiles.
| Topic | Content |
|---|---|
| Calculus | Foundations and advanced topics: limits, derivatives, integrals |
| Linear Algebra | Matrix operations, eigenvalues, spatial transformations |
| Statistical Learning Methods | Full coverage of Li Hang's Statistical Learning Methods 1st edition (Chapters 1–12) |
| Machine Learning | Core content from Zhou Zhihua's Machine Learning |
| Topic | Content |
|---|---|
| Large Language Models | Transformer architecture, LLM training (RLHF), reasoning models, weight quantization, deployment and load testing |
| LLM Engineering | Implementing Qwen3 from scratch, MCP protocol, OpenAI API, various platform APIs, Huawei 910B deployment |
| Deep Learning | PyTorch custom datasets, TensorFlow 2.0, activation functions, MindSpore framework |
| Machine Learning | Sklearn, L2 regularization, hyperparameter tuning (Hyperopt), ensemble learning, time series (Prophet) |
| Mathematics & Information Theory | Information entropy, optimization, GPU FLOPS analysis, LaTeX formulas, reading notes on The Beauty of Mathematics |
| Data Engineering | NumPy, Pandas, PySpark, Elasticsearch, data structures and algorithms |
| DevOps & Tools | Docker, Linux common commands, Python async programming, web scraping, Oracle database |
Continuously updated.