https://github.com/prakharchoudhary/capstone

My capstone project for Udacity Machine Learning Nanodegree
https://github.com/prakharchoudhary/capstone

dataset django jupyter-notebook nltk python scikitlearn-machine-learning sentiment-analysis

Last synced: 20 days ago
JSON representation

My capstone project for Udacity Machine Learning Nanodegree

• Host: GitHub
• URL: https://github.com/prakharchoudhary/capstone
• Owner: prakharchoudhary
• License: mit
• Created: 2017-05-23T00:34:48.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-08-26T00:05:51.000Z (over 7 years ago)
• Last Synced: 2024-11-23T22:03:25.139Z (about 1 month ago)
• Topics: dataset, django, jupyter-notebook, nltk, python, scikitlearn-machine-learning, sentiment-analysis
• Language: Jupyter Notebook
• Size: 4.82 MB
• Stars: 6
• Watchers: 2
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# MLND - Stock Prediction Using Twitter Sentiment Analysis

## Problem Statement

This project will use supervised learning to predict the next day's stock price status(i.e, whether it will higher or lower than current day's closing price).

This project is a python and scikit learn based implementation of the paper: (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.375.4517&rep=rep1&type=pdf)

## Dataset

The dataset includes:

1. The DJIA stock data from June 2009 to December 2009. It has Open, Close, High and Low values for each day.

2. A collection of 350k tweets between the above mentioned daterange, tagged via a self-written sentiment analysis classifier.

## Scripts

Some useful python scripts and jupyter notebook that helped me achieve my goals:

* ```missing_record.py```: fills in interpolated data for missing dates(eg. sundays, holidays etc.) in the DJIA dataset

* ```SentimentAnalysis.ipynb```: Sentiment analysis project to train on kaggleDataset and prepare classifier which can successfully tag the collected tweets. 

### Report

A final report explaining this project and the surrounding problem domain is available as ```report.pdf```.