I have my own trading method which I have used for past 20 years. This is the model that i develop during my free time (my trading not required me to be 7/24, I am not those day traders or 99% losers). The model and concept is from me. The full code is written with me and another person.
The interesting part is the prediction is almost accurate (at least 75%) until the last one hour, but this is what the market gives, nothing to complaint for.
also, fyi, the price prediction is very close to the close price today (< 0.5%), may signal a whipsaw market for tomorrow. This is just what the model predict
What was the original goal of this thread again? It looks like after 100 predictions, your model is not working.
If after a large amount of tris (~150) and if it is still not working, then 'retune' or just scrap it.
That's why I asked what was the goal of this thread. Show the world how good he is for predicting stock price for his self esteem? He doesn't seem to seek for advice or trying to improve by receiving comments and suggestions. He doesn't seem to want to share his knowledge with the ET community. He could just do the same but without telling anyone. The only thing he shows is that his model is not working. But I don't think that anyone here give it a damn. This thread was promising at the beginning. But after 1 month of this....
Totally agree, this thread seems to be about making lots of noise, no discernable content. I was hoping OP would show some interest, some motivation in answering my question re the formation of his model, even just for the sake of enlightening and bringing some journal interest, but alas, OP could hardly bring himself around to volunteering anything. There's another journal like it, 'Diary of an LSE stock trader' where the motivation of the journal imo is not about ET but about hoping to impress in order to get PM's from prospective customers or other vendors. Boring, these type posters usually make it onto my ignore or block list due to lack of content and abundance of time waste.