Machine Learning

by Media Progress LLC

$99.00 $18.99 / month with a 3-day free trial

6 in stock



Machine Learning Guide

The program solves two different problems of machine learning:
• Binary classification answers true or false based on input data.
• Predicting a numeric value gives the answer as a floating-point numeric value based on the input

Formation of training data:

The program for training works with csv files in which the separator is a comma and the divisor of the fractional part is a period. The file must have a specific name namely {name of the currency pair} -data- {step and type of training} .csv examples BTCUSD-data-firstbin.csv BTCUSD-data-secondbin.csv BTCUSD-data-firstreg.csv BTCUSD-data-secondreg .csv. The file must have an internal header which is the name for each column of input data. The first column should store the results for this dataset and should be called Result; all other columns should be called Val_1, Val_2 … Val_99 they store the numerical values ​​of your indicators or anything applicable for the value in the Result column, the total number of columns can be from 2x ( Result, Val_1) up to 100 (Result, Val_1..Val_99) depending on the amount of data supplied. The encoding of the file does not really matter, since this version of the program does not work with the analysis of the mood of textual quotes, or with the addresses of the location of graphic or other information, but takes only numerical values, but I still recommend using utf-8.

Sending input:
The data must be sent to the address and port specified in the program settings and contain only indicator values plus additional last value the last value should contain a number from 1 to 4, depending on which model you are referring to, separated by commas without spaces, the separator of the integer and fractional parts is a dot, the program will return Result as a predicted response

Getting the output:
An adviser who uses in his trade data predicted by the prog will receive a response of the form False, True, 1,1.13890 where the first two values are the answers for the first and second models of binary classification and the third and fourth values are the answers for the first and second models for predicting a numerical value.


URL site – a field for entering the address of the license verification site • License Key – a key issued by the site after purchasing the program. It can be found in the user’s personal account of the site in the API Keys folder. • License Email – the input field of the mailbox to which the account that purchased the program is registered.
Product ID – input field of the program ID. It can be found in your account
Site User in the Keys API folder
Data Dir – a field for entering the address of a folder with samples stored in the files that were stored beforehand. Please note that the program does not create these files or the folder itself, this data is generated by the advisor for your terminal, the program only processes these files and builds a model based on them.
Adress – input field of the IP address on which the program will receive input data for subsequent analysis, and transfer the response with the predicted value.
Port – a field for entering the free port number on which the program will receive input data for subsequent analysis, and transmit a response with the predicted value.
Currency name – field for selecting a currency pair from which the running instance of the program will work.
Column name – input field for the name of the column with calibrated answers. Leave as is, do not change.
Train time – training time for each of the four models. No longer means better, time is chosen experimentally.
Coef 1: input field for dividing the data files into training and test samples.
Save – key for saving settings.


enable / disable workout prediction models


The choice of prediction algorithms as well as metrics for which optimization will be performed for numerical value prediction samples. Similarly for binary classification.

Media Progress LLC

Chat with Vendor
Chat with Vendor
Questions, doubts, problems? We are here to help you!
At the moment, none of our operators is available. Please try again later.
Our operators are busy. Please try again later.
Are you Questions? LET Me KNOW!
This chat session is over
Was this conversation useful? Vote this chat session.
Good Bad