“FIN Status Report 2017-9-4”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(相同用户的一个中间修订版本未显示)
第6行: 第6行:
 
1. Learning with the supervision of the best port does not get good performance yet. And it has large variance.<br>
 
1. Learning with the supervision of the best port does not get good performance yet. And it has large variance.<br>
 
2. Pre-training does not help.<br>  
 
2. Pre-training does not help.<br>  
3. Implement the baseline corn, a pattern-based trading algorithm, with python and improve the time cost from 1.5h+ to 5min-.
+
3. Implement the baseline corn, a pattern-based trading algorithm, with python and reduce the time cost from 1.5h+ to 5min-.
 
[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=wangyang&step=view_request&cvssid=604 results]
 
[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=wangyang&step=view_request&cvssid=604 results]
  
 
||  
 
||  
 
1. Do some experiments and investigate why corn is effective. <br>
 
1. Do some experiments and investigate why corn is effective. <br>
2. Think about some idea to improve corn.
+
2. Think about some simple ideas to improve corn.
  
 
|}
 
|}

2017年9月4日 (一) 02:32的最后版本

People Last Week This Week
Yang Wang

1. Learning with the supervision of the best port does not get good performance yet. And it has large variance.
2. Pre-training does not help.
3. Implement the baseline corn, a pattern-based trading algorithm, with python and reduce the time cost from 1.5h+ to 5min-. results

1. Do some experiments and investigate why corn is effective.
2. Think about some simple ideas to improve corn.