课程将以异常检测模型, 欺诈检测, 推荐系统为导向,从模型的数学理论解释、参数选择、编程技巧三方面进行讲解,完整的带领学生从数据收集与清理到建模输出做两个复杂的数据科学项目。学生可以自己定义感兴趣的问题,并在导师指导下选择数据集,结课时展示自己的完整模型与阐释
1
Introduction to Big Data & Data Analysis
考察:Select an interesting topic and dataset to start a data project
授课1小时; 学习1小时
2
Python & SQL for Data Scientist/Machine Learning Scientist -data manipulation -data visualization -machine learning model
考察:Coding tasks: from easy to hard Project: Use Python/R to clean data and visualize data
授课2小时; 学习4小时
3
Design a anomaly detection beyond prediction and correlation Implement several unsupervised models from end to end
考察:Project: feature engineering, dimension reduction
授课3小时; 学习8小时
4
Fraud detection System & Recommendation System -Design a fraud detection system for attacker prevention and recommendation system for product recommendation
考察:Project: Build a machine learning model from scratch and train - validate - test it from end to end
授课5小时; 学习10小时
5
Evaluate the model, interpret the results and tell a good story
考察:Finish the project and get reviewed and feedback from mentor
授课1小时; 学习2小时