Vectorial model for progressive adaptation for purchase and sale of shares using stock market indicators

Abstract

The shares are considered as a fundamental part of the equity market, as their values change over time as a result of offer and demand, and the effect of market volatility. This volatility makes the trading of shares on a stock exchange is an extremely difficult task. That is why in this article develops and analyzes a system for automatic trading shares, which incorporates a series of progressive learning vector models inspired by the structure of a support vector machine. For the configuration of the overall structure of the model, a number of stock market indicators used by investors to establish positions for buying and selling, were used while learning the model used a sequential negotiation strategy on five different shares listed on the stock exchange of Colombia, and where learning was guided by buying and selling positions that were setting each input stock market indicators. Results from the system showed the profitability that the model was achieved in the negotiation as a result of progress in learning that each of the models was achieved along the sequence of actions used for this study, making the system each more robust time, which makes it ideal for trading shares based on stock indexes.