Google Stock Price Prediction
- 0 Collaborators
To Predict the Future Stock price of Google stocks in share market using the performance of the company over the last 5 years. Deep Learning Model to Predict if the stocks in First month of Jan 2017 will rise or fall,and hence compare with the real performance of the company. ...learn more
Project status: Published/In Market
RealSense™, HPC, Internet of Things, Artificial Intelligence
Overview / Usage
A recurrent neural network (RNN) is a class of artificial neural network where connections between units form a directed graph along a sequence. This allows it to exhibit dynamic temporal behavior for a time sequence.While the training happens we observe the problem of vanishing gradient . To solve this we use LSTMs(long short term memory networks) .Stock Market prediction is a very curious subject where everyone wants to know the future trend/performance of the company.Here we use the concept of memory used in RNN and try to find out the trend by analyzing and understanding the previous performance .
Methodology / Approach
First we use the data set of the google stocks performance of last 5 years.Here i have used data set from year 2011-2016 . After importing all the libraries use import the data set .Now we use feature scaling(mean normalization) to convert all data into similar scale now we create the data set into time steps of 60 (assuming 20 working days in a month) and try to get 1 output ,ie: we take 3 months performance and try to predict the trend for next day of the month and henceforth. Then we build the RNN and visualize the results of the actual performance vs the predicted performance
Technologies Used
Language
● Python
Machine Learning Libraries
● TensorFlow
● Keras
● numpy
● pandas
● MatplotLib