Machine Learning: kNN-based Strategy (mtf)

This is a multi-timeframe version of the kNN-based strategy.
Catatan Rilis: Added custom performance testing.
Catatan Rilis: Updated signal filtering, added data normalization plus some minor fixes.
Catatan Rilis: Price fix in the custom performance testing module.
Catatan Rilis: Minor fix.
Skrip open-source

Sejalan dengan semangat TradingView, penulis skrip ini telah mempublikasikannya secara open-source, sehingga para trader dapat memahami dan memverifikasinya. Salut untuk sang penulis! Anda dapat menggunakannya secara gratis, tetapi penggunaan kembali kode ini dalam publikasi diatur oleh Tata Tertib. Anda dapat memfavoritkannya untuk dapat menggunakannya didalam sebuah chart.

Inggin menggunakan skrip ini pada chart?


how do you change this to a strategy so we can backtest it on the charts?
+6 Jawab
hello, new to this.. can you show results of backtesting? or how to "edit it" to get the result?
+1 Jawab
that was quick luv u bro
+1 Jawab
is there any progress about repainting?
Teach Me Please
Thanks @capissimo . for powerful and strong signals using Hyperbolic Tangent function and Overbought , Oversold levels.
Hi @capissimo, does it repaint? I remember you confirmed it did but I am not sure if it does it anymore.
capissimo dipiers
@dipiers, you should understande the main principle of the algo: it collects data points and selects the ones that are the closest to the current bar's close price. Thus, it repaints. Repainting is yet to be solved. Hopefully, the local community would help.)
+1 Jawab
I really like the backtest results you can see in the study. Could be useful for calculating optimal settings based on profit for each setting.

I also tried turning this into a strategy but the results dont come back as amazing as they look on chart backtest. Should the trades be made as soon as there is a signal or do we still wait for the bar to close? If the close then not sure why the tradingview strategy backtest is so different.
capissimo OrionAlgo
@OrionAlgo, i have the same questions regarding TV backtesting results, so I devised my own empirical calculation as a uniform criteria to compare different strats. Apart from ML-based ones, I use several more indicators to decide about entering a trade.
+1 Jawab
Beranda Penyaring Saham Penyaring Forex Penyaring Crypto Kalender Ekonomi Bagaimana Cara Kerjanya Fitur-Fitur Chart Harga Akun Refer seorang teman Tata Tertib Pusat Bantuan Solusi Website & Broker Widget-Widget Solusi Charting Perpustakaan Charting Ringan Blog & Berita Twitter
Profil Pengaturan Profil Akun dan Penagihan Teman yang di refer Koin Tiket Dukungan Saya Pusat Bantuan Pesan Pribadi Obrolan Keluar