STOCK-MARKET-PREDICTOR

(I) Introduction :

Business and finance sector is today the leader of the world's economy, stock market trading is a major practice in the finance sector. Financial exchange predictions are always trickier when it comes to stock market predictions. It is basically a technique where one tries to predict the future value of current stocks of a company to avoid the loss or perhaps gain profit. This project will demonstrate a machine learning approach to predict the same using various quantities mentioned later in this report. Python is the programming language used for better reach and understanding. We propose a Machine Learning Algorithm which will be trained from different datasets of some companies available from the past to make near effective predictions. Stock market prediction is a technique to determine the upcoming worth of a corporation’s stock or other financial instrument traded on an exchange. A noticeable consequential gain is the sole purpose of stock market prediction, and, of course, to avoid significant losses. Some individuals may disagree with the authenticity of results that these predictions considering the efficient market hypothesis that these predictions cannot be made on the presently available data, thus concluding it as inherently unpredictable. But there are numerous tools and technologies that help to gain future trends’ information, thus resulting in effective profits

(II) Dataset Used :

Data Set of Google, AMD and Tesla was used. Tesla - 2010-2017(predicted for a 20 day period) AMD - 2009-2018(predicted for a 20 day period) Google - 2012-2016(predicted for a 20 day period)

(III) Libraries used :

Numpy :

Fundamental package for scientific computing in Python3, helping us in creating and managing n-dimensional tensors. A vector can be regarded as a 1-D tensor, matrix as 2-D, and so on.

Matplotlib :

A Python3 plotting library used for data visualization.

Tensorflow-Keras :

Is an open source deep learning framework for dataflow and differentiable programming. It’s created and maintained by Google.

Pandas :

Used for data manipulation and analysis. In particular, it offers data structures and operations for manipulating numerical tables and time series.

Sci-kit Learn :

Scikit-learn is a free software machine learning library for the Python programming language. It features various classification, regression and clustering algorithms including support vector machines and many more.

(IV) Website Overview :

Stocks are important to a business because they can help the corporation quickly gain a lot of capital, raise the prestige of the company with the public since people can now invest in the company, and allow the initial investors to sell off shares and earn money on their investments. We provide an efficient solution for easy investment in the stock market so that a layman can also benefit without having prior knowledge of technicalities that a stock market carries. The Website is named as InvestSmart and it helps people invest smartlyFirstly, the number of various examples belonging to each class were identified and plotted.

The website of Technological Stock Market Prediction has on its front page options to go to services, know about the service, read testimonials, go to the home page, contact or stocks.

From the front page, one navigates to the about us page which explains how InvestSmart works in a simplified manner.

Next, the user is directed to various services offered by InvestSmart which are I. Business Consulting II. Stock Market Prediction III. Market Analysis IV. User Monitoring V. Financial Investment VI. Financial Management

As a prototype, the Stock Market Prediction page is ready as of now. With datasets from companies like Amazon, Google & Data Global, the website accurately predicts the stock of these companies using the LSTM method.

On clicking upon a company, In this case an american semiconductor manufacturing company AMD, the user is directed to a page displaying Company information, current stock values and predictions using the LSTM model. The progress includes embedding graphs to the website with better accuracy. The website also predicts the best time to invest and divest in a short interval of days.

The below two pictures show the real stock price and predicted stock price for google along with the best invest & divest periods. It also gives you an example showing how much money you will get if you invest a certain amount in a certain stock.

Tesla, Inc. is an American electric vehicle and clean energy company based in Palo Alto, California. The company specializes in electric vehicle manufacturing, battery energy storage from home to grid scale and, through its acquisition of SolarCity, solar panel and solar roof tile manufacturing.

Testimonials page

This page comprises a form which users need to fill if they want assistance or they have any complaints. It collects users’ data such as Name, contact details, reason of query, etc.

Teams Page

footer section

(V) Results :

Learning based website for Stock Investment :

Users can make their IDs on the website and go through our services section to learn more about Business Consulting, Stock Market Prediction, Market Analysis, User Monitoring, Financial Investment and Financial Management. Thus providing a trustworthy platform to the users to invest their money to achieve their financial goal

Predicting stock values using Machine learning :

The project is implemented LSTM using Keras API of google’s Tensorflow to predict values of stock while training the algorithm on past data. We concentrated on predicting the trend observed in the value of a stock for the next 20 days from the day of prediction. For now we are covering 3 companies which are AMD Inc., Google LLC and Tesla Inc. For this different python libraries were used as stated above in the report

User friendly output :

The most challenging aspect of the project was to generate an efficient output system that is one, user friendly and easily understandable. For this purpose, after our prediction is completed for a certain company, the web page will also display the best possible scenarios that when a user and Invest and Divest to gain maximum profit. This allows the users to clarify what these graphs mean and what challenges are expected to face to avoid losses.

(VI) Conclusions :

We were able to produce results which are consistent with the methodology proposed. Mainy the focus was to integrate the machine learning algorithms with our website for the completion of this project. This provides a powerful tool in the hands of a layman to observe predicted stocks and invest accordingly. A user friendly terminal which shows when to invest and when to divest gives a sense of trust to these predicted values. It is important to note that the results produced are likely to be the trend in the next 20 days. That explains that in between this time period there might be days which are not in resonance with the result but the end result is surely accurate to predict the trend.