Welcome to all!
I'm Drew, one of the co-founders of MLFX a Forex algorithm based on 10 years of university-level research by my business partner and friend Amaury Hernandez. Our algorithm (Hermes) uses multi-agent systems, intuitionistic fuzzy logic, and bio-inspired machine learning to predict and trade the Forex markets. If that made you scratch your head (trust me you're not alone) then let me break it down for you.
essentially each "AGENT" can be thought of as an expert human trader
These agents each have a set number of unique INPUTS (such as RSI)
Those inputs have RULES assigned to them (if RSI is above 80 do xyz)
a GROUP of RULES is essentially a trading strategy
fuzzy logic lets agents MEASURE HOW TRUE/FALSE each rule is, as opposed to binary true/false
The machine learning aspect iterates through thousands of possible combinations looking for optimally configured agents
If you're still struggling, think Quantopian but instead of human quants creating algorithmic trading strategies, machine learning is creating those strategies instead.
June 2021 Update:
We've decided to shift our focus onto creating something lots of people can use right now, the problem with what we were previously working towards is that it was uncertain when we would have a product that was actually useful to people. Yes we had metrics in mind that we were tracking, but it could have taken us years to get there and we need revenue much sooner than that. So, what are we working on now?
MLFX MVP (minimum viable product):
We're creating a library of popular/requested Forex strategies that can be easily compared and ranked against each other based on historical and live performance. Strategies become inputs for our machine learning agents who iteratively find better configurations of that strategy.
As an MLFX user, you can follow any of these strategies, plot historical and live trades on our charts, and receive live notifications when new signals are created.
WHY?
Reduce time spent scanning charts for setups that meet your criteria
Giving you more time to focus on trade execution and which setups to trade
Resulting in less trader fatigue and better trading decisions
Reduce uncertainty in the performance of your strategy
Giving you the confidence to stick with a winning strategy through a rough period
Resulting in more consistent long-term trading results
We're currently looking for an experienced *profitable* trader to join our team in an advisory role - You would be assisting with the creation of our products to make them more trader friendly and advising us on areas we should priorities when it comes to algorithm development. For instance, indicator X should be focussed on before indicator Y. Send me a message if you're interested.
Our Telegram group is a great way to keep in contact with us: https://t.me/joinchat/IFyZTEw_uhEHe0Oo5uSAVw
REQUEST A STRATEGY IN THIS THREAD:
If you have a strategy that you want us to add to MLFX then request it in this thread and we'll let you know when it has been added. This will eventually be a paid feature, something like $50-$100 per strategy, but as long as you're seeing this then it's free!
MLFX RESEARCH:
All research found here:
Using Fuzzy Inference Systems for Creation of Forex Market Predictive Models
MONTHLY UPDATE ARTICLES:
July 2020: | Launch of Analysis Page
August 2020 | Implementation of Multi-Objective Algorithm
September 2020 | A Pattern Thingy
October 2020 | 200 THOUSAND Times Faster
November 2020 | Storing and Plotting Data
December 2020 | Manual Dataset Selection & The Mactavator
January 2021 | Bug Squashing Time
February 2021 | The Team Grows
March 2021 | Nested Signals
April 2021 | A Beautiful Vue
May 2021 | Eating Humble Pie
I'm Drew, one of the co-founders of MLFX a Forex algorithm based on 10 years of university-level research by my business partner and friend Amaury Hernandez. Our algorithm (Hermes) uses multi-agent systems, intuitionistic fuzzy logic, and bio-inspired machine learning to predict and trade the Forex markets. If that made you scratch your head (trust me you're not alone) then let me break it down for you.
essentially each "AGENT" can be thought of as an expert human trader
These agents each have a set number of unique INPUTS (such as RSI)
Those inputs have RULES assigned to them (if RSI is above 80 do xyz)
a GROUP of RULES is essentially a trading strategy
fuzzy logic lets agents MEASURE HOW TRUE/FALSE each rule is, as opposed to binary true/false
The machine learning aspect iterates through thousands of possible combinations looking for optimally configured agents
If you're still struggling, think Quantopian but instead of human quants creating algorithmic trading strategies, machine learning is creating those strategies instead.
June 2021 Update:
We've decided to shift our focus onto creating something lots of people can use right now, the problem with what we were previously working towards is that it was uncertain when we would have a product that was actually useful to people. Yes we had metrics in mind that we were tracking, but it could have taken us years to get there and we need revenue much sooner than that. So, what are we working on now?
MLFX MVP (minimum viable product):
We're creating a library of popular/requested Forex strategies that can be easily compared and ranked against each other based on historical and live performance. Strategies become inputs for our machine learning agents who iteratively find better configurations of that strategy.
As an MLFX user, you can follow any of these strategies, plot historical and live trades on our charts, and receive live notifications when new signals are created.
WHY?
Reduce time spent scanning charts for setups that meet your criteria
Giving you more time to focus on trade execution and which setups to trade
Resulting in less trader fatigue and better trading decisions
Reduce uncertainty in the performance of your strategy
Giving you the confidence to stick with a winning strategy through a rough period
Resulting in more consistent long-term trading results
We're currently looking for an experienced *profitable* trader to join our team in an advisory role - You would be assisting with the creation of our products to make them more trader friendly and advising us on areas we should priorities when it comes to algorithm development. For instance, indicator X should be focussed on before indicator Y. Send me a message if you're interested.
Our Telegram group is a great way to keep in contact with us: https://t.me/joinchat/IFyZTEw_uhEHe0Oo5uSAVw
REQUEST A STRATEGY IN THIS THREAD:
If you have a strategy that you want us to add to MLFX then request it in this thread and we'll let you know when it has been added. This will eventually be a paid feature, something like $50-$100 per strategy, but as long as you're seeing this then it's free!
MLFX RESEARCH:
All research found here:
Using Fuzzy Inference Systems for Creation of Forex Market Predictive Models
MONTHLY UPDATE ARTICLES:
July 2020: | Launch of Analysis Page
August 2020 | Implementation of Multi-Objective Algorithm
September 2020 | A Pattern Thingy
October 2020 | 200 THOUSAND Times Faster
November 2020 | Storing and Plotting Data
December 2020 | Manual Dataset Selection & The Mactavator
January 2021 | Bug Squashing Time
February 2021 | The Team Grows
March 2021 | Nested Signals
April 2021 | A Beautiful Vue
May 2021 | Eating Humble Pie