This project aims to predict future Bitcoin prices relative to USD by analyzing historical price data. It involves data preprocessing, exploratory data analysis, statistical testing, and the application of machine learning algorithms. The project demonstrates proficiency in essential data science tools such as pandas, matplotlib, seaborn, and scikit-learn.
-
Data Collection: Eight years of historical Bitcoin-USD price data were collected for analysis.
-
Data Preprocessing: The collected data was preprocessed and cleaned to ensure quality and consistency for analysis.
-
Exploratory Data Analysis: Visualization techniques were used to identify patterns, trends, and potential outliers in the dataset.
-
Statistical Testing: Assumptions and relationships within the data were validated through statistical testing.
-
Machine Learning: Decision tree algorithms were applied and tuned to predict Bitcoin's opening price.
-
Model Evaluation: The performance of the model was evaluated, identifying areas for potential improvement and discussing possible extensions of the project.
The tuned decision tree predicted a Bitcoin opening price of 34,744.66 USD, which was close to the actual price of 34,225.67 USD.
- google-colab
- pandas
- numpy
- matplotlib
- os
- datetime
- scipy
- seaborn
- scikit-learn
To install the required dependencies, run: