Python course
CSX433.3 Python for Data Analysis and Scientific Computing CSX433.7 Machine Learning with Tensorflow
Final Project -- Bitcoin Cryptocurrency Price Prediction using RNN, LSTM with Tensorflow
Alison Jing Huang -- jing01ucsb@gmail.com Github: https://github.com/alisonjing
Elaine Phan -- elainph@gmail.com Github: https://github.com/ElainePhan
The purpose of this project is to create a basic economic prediction model to estimate the future trend of bitcoin. The presentation will include functions to input bitcoin purchases at different prices from hour to hour, day to day, month to month and year to year. The data is loaded in xlsx text file format with different parameters such as purchase dates, purchase prices, open prices, high prices, low prices, purchase volume, purchase volume from, purchase volume to, time series and percent changes of price during a period of time.
We plan to analyze the bitcoin data looking at low and high prices per day, month, and year using NumPy arrays, Matplotlib, and Scipy. We will use these libraries to analyzed the data to train our models with analytical tools such as Linear Regression, Time Series, Machine Learning (Long Short Term Memory (LSTM)) and Tensorflow. We will test our models using these tools to make prediction for the future trends of bitcoin.
Data Source: https://www.investing.com/crypto/bitcoin/historical-data https://www.cryptodatadownload.com/index.html
The code was created in Anaconda Jupyter Notebook enviroment(version 5.0.0)
The following modules/packages are required, and can be installed using pip: numpy scipy datetime pandas pandas-datareader matplotlib matplotlib.pyplot pylab sklearn statsmodels seaborn tensorflow kera ARIMA