文件名称：portfolio-manager:使用强化学习来管理产品组合，使用DQN，Q-Learning和DDPG模型
文件大小：1.21MB
文件格式：ZIP
更新时间：2024-06-04 13:34:13
Python
一，包装简介： |-- State_Space: this package is used to generate daily state space and portfolio IR rank. | |-- feat_selection.py: Functions mainly used to select features. | |-- Generate_IR_rank_week.py: Generate information ratio rank based on weekly data. | |-- Update_IR_rank.py: Functions used to update IR rank every week. | |-- data: this folder contains required data. | |-- industry.csv :
【文件预览】：
portfolio-manager-develop
----task2()
--------DQNmodel()
--------Processor()
--------utils.py(1KB)
--------QLearningModel_beta()
--------QLearningModel()
--------__init__.py(0B)
--------DDPGmodel()
--------store()
----data()
--------industry.csv(61KB)
--------IR_rank_week.csv(388KB)
--------tradingday.csv(27KB)
----task1()
--------Supervised_Learning()
--------Similarity_Search()
--------__init__.py(0B)
----result()
--------task2_result.csv(33B)
--------.gitkeeper(0B)
--------report.docx(773KB)
----__init__.py(0B)
----run_task2.py(2KB)
----.gitignore(184B)
----State_Space()
--------IR_month.py(4KB)
--------feat_selection.py(12KB)
--------Update_IR_rank.py(8KB)
--------IR_rank_month.csv(151KB)
--------__init__.py(0B)
--------IR_rank_week.csv(388KB)
--------Generate_IR_rank_week.py(5KB)
----README.md(3KB)
----run_task1.py(253B)
----base.py(2KB)