机器学习导论 - Introduction to Machine Learning

课程介绍 Course Introduction

学分：4 | 先修课：线性代数、概率论与数理统计、程序设计基础 | 学期：第4学期

机器学习导论系统介绍机器学习的基本概念、经典算法与理论基础。课程涵盖监督学习（线性回归、逻辑回归、决策树、支持向量机、集成学习）、无监督学习（聚类、降维、密度估计）以及模型评估、正则化、偏差-方差权衡等内容。学生将通过编程作业掌握算法实现，理解过拟合与泛化，为深入研究深度学习与统计学习打下基础。

Introduction to Machine Learning systematically covers fundamental concepts, classic algorithms, and theoretical foundations. Topics include supervised learning (linear and logistic regression, decision trees, support vector machines, ensemble methods), unsupervised learning (clustering, dimensionality reduction, density estimation), model evaluation, regularization, and the bias-variance trade-off. Programming assignments build implementation skills and intuition for generalization, preparing students for deep learning and statistical learning.

大作业 Final Project

作业标题：房价预测回归系统

学生需基于真实房价数据集（如Ames Housing）构建回归预测系统。要求完成数据清洗、特征工程、多模型对比（线性回归、随机森林、XGBoost等）、交叉验证与超参数调优，并分析特征重要性与误差来源。需提交可运行代码、技术报告及预测结果排名，RMSE作为主要评价指标。

Students build a regression system on a real housing dataset such as Ames Housing. The project requires data cleaning, feature engineering, multi-model comparison (linear regression, random forest, XGBoost), cross-validation, hyperparameter tuning, and analysis of feature importance and error sources. Deliverables include runnable code, a technical report, and leaderboard predictions evaluated by RMSE.