Thanks for the awesome and educational content. I am backtesting a strategy that follows “secondary day” where I expect price to keep going up after a strong day that closed close to the highs. My stop loss is well defined but I am struggling to compute a good take profit, and because of it I am seeing only 43% win ratio. Do you have any advice for me?
@MLAlgoTrader4 ай бұрын
I am glad you find it interesting. I think you would need to maybe optimize this parameter with backtesting. I will say this will be "overfitting", but it can give you some ideas. Soon I will make backtesting videos. The smaller the take profit the higher the win ratio. Similarly, the higher the stop loss the higher the win ratio, but when you lose you lose a lot more. I also want to say I am not a momentum guy and more of a dip person so I like when stocks go down, but this has cost me a lot of money in real life :)
@JustFlow-it2zg4 ай бұрын
for Future request, please consider 15min Opening range breakout strategy , also optionally take decision based on 50SMA above or below and take trade in that direction only
@MLAlgoTrader4 ай бұрын
Yes I got that request haven't got to it
@peterk62154 ай бұрын
Hi. Can you upload again at higher quality please? It's only 360p.
@MLAlgoTrader4 ай бұрын
Ehh I spent a few hours trying to fix the video lol. It might just be processing try back in a bit. I had issues combining the video for some reason lol. Video editing is hard!
@MLAlgoTrader4 ай бұрын
1080 seems to work now lol .
@am7-p4 ай бұрын
It takes time for YT encoders to finish their jobs.
@paulrelf4 ай бұрын
Very nice to see an end to end example. Thanks for that!
@MLAlgoTrader4 ай бұрын
Thanks always happy to see your comments.
@אוריינקוביץ-ש8צ4 ай бұрын
I see a algo but no ML.
@MLAlgoTrader4 ай бұрын
There is an algorithm called Dummy Classifier. I explain in my other video how you learn an algorithm. I assume you already have a a model learned in this video.. Here I explain one possible way to learn a model but you need your own kzbin.info/www/bejne/oqS7nXh3qbCIprcsi=3FXTBYhRiMwSoc65
@MLAlgoTrader4 ай бұрын
I wanted to add something. I purposely kept the algorithm "bad/basic" This is an algorithm you should do your own research on. I just showed the minimal idea. ML for algotrading is not ML for weather prediction. The signal to noise is a lot worse so you have to be ultra careful. I can try to make a video showing a possible algorthmic improvement say using (MSFT, GOOG,GOOGL,AMZN,META,TSLA,NVDA) + their weightings + relative volumes as features to predict if QQQ gap will close, but even if I show that you need to assume my algorithm is WRONG!
@am7-p4 ай бұрын
We we have access to the notebook source code by any chance?
@MLAlgoTrader4 ай бұрын
I don't feel comfortable releasing it yet but will once I double check a few things .