PhD Tested

Neural Network Technology Used to Predict Market Trends

Neural Networks – The Mathematics behind Intermarket Analysis

Neural Network
Map of a successful neural network trading program. VantagePoint is an example of an analytical software program that uses multiple neural networks to analyze data and produce market forecasts.

Mr. Mendelsohn has spent millions of dollars and more than two decades researching and applying neural networks to the global financial markets. The result has been nothing short of phenomenal. He and his research team, the Predictive Technologies Group, have successfully developed and refined an extremely sophisticated, proprietary, patented technology that implements this entire process.

This includes the identification and selection of the most highly related intermarkets, choice of data, preprocessing of the data, training of the neural networks to make future forecasts through proprietary leading indicators, detecting forecasting errors, and having the neural networks make necessary adjustments through an interactive training process.  Then, independent (out-of-sample) data is used to evaluate the forecast predictions under real-time conditions, compare performance results from thousands of different neural networks, and determine which neural network configuration to use in the final application to maximize the predictive accuracy at forecasting the short term trend directions of hundreds of global markets each day.

Artificial Intelligence

If you would like to learn more about Mr. Mendelsohn’s technical breakthroughs in applying neural networks to the global financial markets and technical analysis, click here to see an ARTIFICIAL INTELLIGENCE Special report.

Neural Networks
Neural networks continuously try to find hidden patterns. like the human brain, neural networks “learn” by sifting through data over and over again to find patterns.

In a nutshell, artificial neural networks are obviously not as complex as a human brain, but can assume brain-like functions as they study data, “learn” subtle relationships within and between related markets, recognize hidden repeating patterns in global market data, and use this information to make highly accurate predictive market forecasts. Similar to the human learning process, artificial neural networks can be designed to learn patterns from exposure to repeated examples of those patterns. This repeated exposure allows neural networks to generalize conclusions about related but previously unseen pat­terns. This process is what gives artificial neural networks their ability to make forecasts of future market patterns such as the future short term trend direction of individual markets.

Mr. Mendelsohn has written extensively about neural networks and their application to the global markets and trend forecasting in numerous books and dozens of technical articles in financial journals over the past two decades.  To see Mr. Mendelsohn’s collected works specifically on the topic of Neural Networks, go here >


FREE Support
& Training

Vantage Point Awards
& Software Reviews

Free Recent
Forecasts & Live Demo

86% Challenge
The Results Are In