What Makes A.I. for Trading
Such a Challenge?

If you explore the internet, you will find many claims of A.I. Beating the market.  But most of them are  simply Trading Systems  based on techniques that have been around for over 50 years.  Training any A.I. to generate consistent profits in the market is difficult.

Many Artificial Intelligence Models are based on a technique called "Classification."  This type of Model combines a number of Features in an attempt to classify a situation or object. They do well at things like  identifying  a cat or a dog in an image,  or predicting events from sales data, The problem is, the market is more like a snowflake - no two market days are exactly alike. 

In fact, the Stock Market is a highly complex interaction of events and emotions that is always changing.  No two charts are alike, and similar-looking charts can have opposite behaviors on the right edge.   Solving the Market Problem using A.I. requires investigation into new Models PLUS the development of advanced "Features" that indicate what the next move is likely to be.

For a Machine to Beat the Market
It Has to Think Like a Trader.

The most successful Traders look for specific clues in charts that indicate the psychology of the market, including  complex Price  and Volume Patterns.

There are also clues in other timeframes (weekly, for example) and other data, including fundamentals and what the market is doing as a whole.   

The trader's brain combines all this information into an idea of future price movement.  In the end, there is no "solution" - it's all about probability.  And the best traders know when to take profits and losses, based on trading experience.

Super Inputs

The data fed to an A.I. model should  identify high level "Features" like key chart patterns, consensus of movement and other indications of bullish or bearish sentiment.

We call these advanced algorithmic Features,  Super Inputs.  It's the model's job to combine Super Inputs to create profitable Trading Rules.  

Profitable Models have already been created using the Genetic Algorithm.*  Strategies based on them are released to Members periodically.  With the right Super Inputs, we can apply them in the most modern Cloud A.I. Models to provide stock predictions from the Cloud to Members, as explained on the Overview.

The TRADEnosis Group

TRADEnosis was the term coined by A.I. Researcher Steve Mayo when he built his first Cloud A.I. system in 2022.  

We have taken the lessons learned from this early version to pursue the creation and testing of Super Inputs in our Rule-Based Models.  

As we do this, Members of the TRADEnosis Group have access to the interim Strategies. Simply download and run the Strategies in OmniTrader (which we provide if you don't have it.)

Our goal is to deploy TRADEnosis 2 into the Cloud, so it can provide Signals from the market to Members on multiple platforms.