Thank you. I found that portfolio backtesting with multiple assets is more difficult to overfit compared to single asset backtesting. Also, this strategy was created in 2018. However, validation is certainly a great and necessary process during strategy development.
老師真的很厲害
謝謝！我也是邊做邊學哇哈哈！一起加油～
這邊也想請問一些問題，煩請大大看到後能幫忙解惑！
1. 想請問 hold_until() 這個 method 的 spec 可以在哪邊看到，且出來的 output 會是什麼 data structure?
2. 如果想要針對找出的股票清單，要怎麼做 filter(ex 像是只想列出 8 月進場的股票)，或是哪邊有相關的 spec 可以看呢？
1. 可以在這 doc.finlab.tw/reference/dataframe/#finlab.dataframe.FinlabDataFrame.hold_until 找到喔！
2. 最後產生的 report 當中的 report.get_trades() 可以跑出所有的進出場訊號，然後可以用 pandas 的功能去 filter 就可以囉！
It seems to me that you can always get a decent backtesting performance if the cross-validation is not adopted. Other than that, great video though.
Thank you. I found that portfolio backtesting with multiple assets is more difficult to overfit compared to single asset backtesting. Also, this strategy was created in 2018. However, validation is certainly a great and necessary process during strategy development.