Analyzed an American stock exchange, the National Association of Securities Dealers Automated Quotations, better known as NASDAQ.
It is the largest electronic screen-based equity securities trading market in the United States.
Used ARIMA + GARCH model and machine learning techniques Naive Bayes and Decision tree to determine if we go long or short for a given stock on a particular day.
There are two parts in the project:
(a) ARIMA + GARCH model helps us to predict the overall trend for the NASDAQ 100 index. On the basis of the observations from the NASDAQ curve, we then focus on specific stocks to ascertain profitability by going long on them or not.
(b) We implement machine learning techniques like Naïve Bayes and Decision trees to predict if or not the given stock is going to perform positively on the following day given the data for today and in the past.