1. Home
  2. Docs
  3. DLPAL LS
  4. Introduction

Introduction

Deep Learning Price Action Lab for long/short equity (DLPAL LS) uses primitive attributes of price action, and specifically the open, high, low and close, to extract features types in an unsupervised learning mode based on general feature clusters. Then, the program uses the extracted features in supervised learning mode to identify long and short candidates in a universe of securities: The long/short identification is based on a set of calculated features and the user has flexibility in ranking the results according to their values. The output of the program can be saved in csv format and further used in conjunction with trading platforms to automatically execute trades.

DLPAL LS offers capability of generating historical files with the calculated features for all securities in a universe. These files can be used to backtest various long/short equity strategies in third party backtesting platforms. The program also generates historical files with binary targets based on the next day return and scoring files to be used with machine learning algorithms. Therefore, the output of the program can be used to implement both fixed algos based on the calculated features and also algos derived from machine learning classification. The program can also be used for discretionary trading. 

The functions of DLPAL LS are outlined below.

Backtesting and machine learning scoring are done in third-party platforms. DLPAL LS provides all the files needed for those tasks. 

DLPAL v1.0  LS was released March 8, 2017

DLPAL v2.0  LS was released October 1, 2017

– Added returns calculations
– Several improvements
– Updated manual

DLPAL v2.0 LS updated February 5, 2018

– Added next bar returns calculations for close to close

DLPAL v2.0 LS updated March 8, 2018

– Test Files update: Checks for files with invalid delimiters and lists only files with zero values

DLPAL v3.0  LS was released April 8, 2018

– Added filtering by top N/bottom N results
– Added option to set minimum significance S to 1 in results and feature generation
– Several changes to program look and menus

DLPAL  v4.0  LS was released October 8, 2018

– Fixed return calculations in Returns for short positions
– Improved algorithm speed
– Other minor fixes

DLPAL  v4.0  LS update was released June 3, 2019

– Multiple instance option is now a standard feature

DLPAL v5.0 LS was released October 8, 2019

– Update of history files is now possible from a workspace provided each file is listed separately in each input line. This allows bulk update of files with different reward/risk and other parameters
– Other minor fixes

Smoothed features add-on released May 8, 2020

– The add-on offers option for generating smooth features (current and historical.)
– The add-on allows a toggle between all long/all short signals in the results.
– Additional details added for next bar return calculations.

DLPAL LS v6.0 released September 8, 2020

– Minor fixes and additions
– Smoothed features integration

DLPAL LS v6.0  features autoupdate released October 21, 2020

New capability of updating automatically historical features files. Click here for details. 

DLPAL LS v6.0  Execution Automation Add-on released November 27 , 2020

The Execution Automation Add-on  (EAA) allows running the program in the background, automatically saving results to a directory of choice and terminating execution after the run is completed. Click here for more details. 

DLPAL LS v7.0 released March 10, 2022

-Execution Automation Add-on is standard feature.
-Additional options for sorting the results.
-Several minor fixes.