TO PREDICT ACCURATE SILVER MCX TREND IN INDIAN STOCK MARKET USING BOX-JENKINS METHOD

Abstract

In today’s world we are surrounded by loads of data. Data in form of text, images, videos, locations etc. We are moving into a new era that is termed as the Data Era. Data is going to play a vital part by changing the way people interact with their surroundings on daily basis. Experts says Data will lead to the fourth industrial revolution. There are around 2.5 quintillion bytes of data that is being gathered currently each day. This a lot of data and the potential of benefits from this data is unimaginable. Social media platforms like Facebook, Twitter, and Google are one of the major contributors of data in towards world. Today we are well equipped with the technology to deal with this amount of data and use the same for our advantage. Technologies like Machine Learning, Deep Learning, Artificial Intelligence, Block Chain, and 5G provide us with a great opportunity to use in almost each and every field. One of the advantages that data could provide is in prediction of future demands of products, stock market prices, investment plans and market requirements. This could be beneficial to both product/ service providers and the users who avail them. Many big organizations have started using data for their advantage. With the growing competition and demand for better products and services, companies are seeking helping of Machine Learning algorithms to learn the shopping patterns of the customers, get more insight into the department of their company lacking behind and many more such information. As we know today a lot of people are interested in buying stocks and commodities from the market due the promising returns these tools present to the customer. So the price of such tools changes day to day. Every day the stock and commodity have a different price tag attached to it. So there is a series of values associated with the stock or commodity over a period of time. Such process can be termed as Time-Series process in the language of Data Science. Here we are taking up the task of finding the patterns in Time-Series processes so that the future value of such process can be predicted with utmost accuracy. To find the patterns and predict the Page iv

future values will be using the Box-Jenkins approach. This approach has proved to be very accurate for finding patterns and predicting the future values of the Time-Series processes. The language used to apply the Box-Jenkins approach here is python. Python provides various libraries to efficiently work with various machine learning algorithms. The few libraries used here are pandas, numpy, and matplotlib.