Design a fuzzy expert system for stock trading

Design a fuzzy expert system for stock trading, The inputs to the fuzzy system are the current price of a company’s stock, and an indicator called MAD (moving average divergence). The fuzzy sets for the stock price are VL (very low) LO (low), MD (medium) and HI (high) and VH (very high). Similarly the fuzzy sets for the MAD are N (negative), Z (zero) and P (positive). The outputs are BM (buy many), BF (buy few), DT (do not trade), SF (sell few) and SM (sell many). Suppose that there is a correlation between the price and the MAD indicator such that when MAD is positive today the stock price is expected to rise the next day with a highly probability, and when MAD is negative, the stock price is expected to fall the next day with a high probability. Finally when the indicator does not show a noticeable trend, the next day price has no correlation with the MAD. However, there are sometimes exceptions to above rules due to political events and psychological mood of the market (e.g. a controversial president’s tweet, political situation and new from Afghanistan, U.S., Russia and China, corporate scandals and fraud, Homeland Security Department raising the alarm level, etc.), change in earning forecast, etc. The news about these events often produces random fluctuations in stock prices, which we will model as random numbers. Thus the stock share price p(i) and MAD denoted by m(i), at the open of the i-th day trading are modeled as P(i)=10+2.8 sin⁡(2πi/19)+0.9 cos⁡(2πi/19)+ζ(i) m(i)=0.6 cos(0.4i)-sin⁡(0.5i)+η(i) where i=1,2,...,150 is the day number, ζ(i)=0.6r(i) ×(i%3) , η(i)=0.5r(i) ×(i%3), and r(i) is a random number between –1 and +1. Design a fuzzy trading program to maximize your profit (or minimize your loss). Note that the fuzzy system can only find today’s price and today’s MAD and has no knowledge of future prices or future MAD given by the above equations. Simulate the system assuming that you can trade no more than once a day at the market open. You have $10,000 to invest, and the maximum number of stock share buy or sell each time is 600. Develop your own program for fuzzification, rule evaluation, defuzzification, etc.