人工智能与机器学习

groupingCOSMOMargulesall.ipynb 詳細分析報告

本 Jupyter Notebook 實作了一套物理約束神經網路 PINN 預測管線，用於預測氣液平衡 VLE 數據。方法結合了 COSMO 分子描述符與 Margules 活度係數模型。Notebook 載入預先訓練好的模型，針對二元混合物的熱力學性質（氣相組成 y 及系統壓力 P）進行預測，共約 17,600 筆數據。

AI for Chemistry 專案原理與運作說明

1. 專案概述專案概述 2. 基礎理論基礎理論 3. 系統架構系統架構 4. 資料處理流程資料處理流程 5. 模型架構詳解模型架構詳解 6. 訓練與預測流程訓練與預測流程 7. 技術細節技術細節

RAG + AI 完整教材

從零開始打造一套生產級 RetrievalAugmented Generation（檢索增強生成）系統 主要技術棧：LangChain + OpenAI + Chroma；輔以 Ollama + FAISS 本地化方案 適合對象：Python 開發者、AI 應用工程師、資料工程師

RAG AI 範例專案

依據 RAGAI完整教材.mdRAGAI完整教材.md 附錄 A 實作的生產級 PDF / Markdown 智能問答系統。

范例科技 — 常見問題 FAQ

超過 rate limit 會回傳 HTTP 429，建議使用指數退避（exponential backoff）重試。

范例科技股份有限公司員工手冊

SSD — Sitio Web Portfolio

核心目标： 在大四实习期间作为求职工具，向西班牙设计公司展示个人能力 传达独特的设计哲学："轻足迹"空间设计理念 毕业后作为持续更新的职业名片

Chijia · Portfolio personal

Diseño donde los humanos aterrizan suavemente. 设计，让人轻轻落脚的地方。

Detailed Analysis of groupingCOSMOMargulesall.ipynb

This Jupyter notebook implements a PhysicsInformed Neural Network PINN prediction pipeline for VaporLiquid Equilibrium VLE data using COSMObased molecular descriptors combined with the Margules activity coefficient model

chempropVP

Note: These environments include a CPUonly version of PyTorch. If GPU acceleration is desired, please refer to the PyTorch homepagehttps://pytorch.org for the correct version and installation command. The recommended PyT

AI for Chemistry 問題分析與改良方案

1. 程式碼品質問題程式碼品質問題 2. 架構設計問題架構設計問題 3. 資料處理問題資料處理問題 4. 模型設計問題模型設計問題 5. 訓練流程問題訓練流程問題 6. 可擴展性問題可擴展性問題 7. 綜合改良建議綜合改良建議

Software Requirements Specification SRS

Project: Pediatric Q&A via LINE Python Flask + LINE + RAG with OpenAI

Comprehensive Machine Learning & Deep Learning Curriculum

A complete set of professional, textbookquality teaching PDFs covering a depthfirst path from mathematical foundations to production deep learning — written for someone who already codes well but wants real understanding

How to use this curriculum

Machine Learning Deep LearningA depth-first curriculum, from math foundations to production9–14 months at 8–12 hrs/weekTurn ``I can call '' into ``I understand, can implement, and can ship.'' This guide is the map for th

Preface: how to use this book

There is one equation behind this entire book, and it fits on a single line: [ P( D)=P(D),. ] is a recipe for learning from evidence: start with what you believe (the ), see how well each possibility explains the data (t

Distributions, Conjugate Pairs, and Formulas

1.5 @p4.8cmp9.2cm@ Bayes' theorem P( D)=P(D) Working form p( D) p(D),p() Evidence (marginal likelihood) p(D)= p(D),p(),d Posterior predictive p( x D)= p( x),p( D),d MAP estimate =_ Bayes action a=_a _ D[L(,a)] Bayes fact

What Is Bayesian Thinking?

Before any formula, Bayesian inference is a : hold beliefs as probabilities, and revise them in proportion to the evidence. This chapter sets up that worldview — probability as degree of belief, the contrast with the fre

Bayes' Theorem

Now we make the update rule exact. is a short consequence of the definition of conditional probability, but its implications are deep enough to have reshaped statistics, science, and AI. This chapter derives it, names it

The Prior, the Likelihood, and the Posterior

In Chapters– the hypotheses were discrete (sick or healthy). The real power of Bayesian inference appears when the unknown is a — a probability, a rate, a mean — so the prior and posterior become . This chapter is the co

Conjugate Priors and Updating

Computing the posterior means doing the integral p(D)= p(D)p(),d — usually hard. are a beautiful special case where the integral is free: prior and posterior belong to the same family, so updating is just . The three cla

Choosing Priors

The prior is what makes Bayesian inference both powerful and controversial: it lets you inject knowledge, but it also forces you to be explicit about your assumptions. This chapter is a practical guide to choosing priors

Summarizing the Posterior and Bayesian Decisions

The posterior is a full distribution, but people need answers: a single estimate, an interval, a yes/no decision. This chapter shows how to a posterior responsibly (point estimates and credible intervals) and how Bayesia

Naive Bayes Classifiers

is Bayes' theorem turned into a fast, surprisingly effective classifier. Despite an assumption that is almost always false — that features are independent given the class — it powers spam filters, document classifiers, a

Bayesian versus Frequentist Inference

The two great schools of statistics answer the same questions in fundamentally different ways. Understanding the contrast sharpens your grasp of — and clears up the most persistent confusions in applied statistics, like