项目跨度

  3 – 4 个月

项目时长

  约150 小时

上课频次

  9 – 12 小时/周

项目进行方式

  实战为主,理论为辅

工业级项目介绍

   中级 – Machine Learning Capstone Project

本项目旨在让同学们通过经典数据集,熟练掌握

  • Data wrangling using pandas data frame
  • Machine learning methods including supervised and unsupervised learning
  • Analyze and visualize model results

   中级 – Spark Recommendation System Project

通过完成此推荐系统课题,同学们不仅能够掌握该项算法,而且可以学习在大数据领域最经典的一款工具Spark

   进阶 – Deep Learning 无人超市商品识别Project

本项目引导同学在熟练掌握Deep Learning技术的同时,将该技术(尤其是Computer Vision)应用到现实生活中。我们将与国际知名食品公司Easy Way Group深度合作,为其开发一道商品识别系统

   进阶 – Capital Market & Reinforcement Learning Project

— 项目中也将讲授资本市场的框架结构,对比介绍基本金融产品(股票,债券,基金,期权等)的特点和风险。结合Monte Carlo模拟(Python),引入量化分析金融产品以及投资组合管理的常识性概念和方法,并实际应用强化学习来优化股票交易流程及指导操作

适合人群

   任何背景的学生都可以从Python Fundamental学起;如之前已有Python, Machine Learning的基础, 可从中级-Data Science Industry Project做起

对应岗位

   Data Scientist, Machine Learning Engineer, AI specialist, Data Analyst, Business Analyst, Data Engineer, Software Engineer

流程图
项目大纲

初级-Basic & Advanced Python | 约 30 小时

  Python Basics 
  • Github basic and anaconda setup
  • Python basic: variable, list, dictionary, the basic calculation
  • Python intermediate: Class
  • Wrap up and exam

  Data Frame

  • Basic syntax
  • Data wrangling
  • Data exploration and analytics
  • Wrap up and exam

中级-Data Science Industry Project | 约 40 小时

  Machine Learning

  • Statistics basics
  • Linear regression
  • Feature engineering
  • Tree-based model
  • Model interpretability
  • Unsupervised: k-means + hierarchical linkage / DBSCAN
  • Wrap up and exam

  Spark Recommendation

  • Spark setup and basics
  • Recommender system
  • ALS and model evaluation
  • Project finish up

进阶-Data Science Premium Industry Project | 约 80 小时

  Object Recognition

  • Neural network and CNN
  • Transfer learning
  • Data collection and model building
  • Model improvement

  Capital Market

  • Capital market basics
  • Reinforcement learning
  • Data collection and investigation
  • Model design and implementation