(Built as part of INFO 52272 - Data Science using python in Summer'18 under guidance of Prof. Dino)
🌎 The project analyses trading statistics of goods across nations with tools like- Bayesian model, statistical hypothesis testing, ANOVA and feed forward neural network along with visualisations! 🌎
💰 💰 💰
The United Nations Commodity Trade Statistics Database (UN Comtrade) complete trade records between countries and publishes it for free. We have fetched data from 1985 to 2016 which is ~1.2 Gb in size. Over 140 reporter countries provide the United Nations Statistics Division with their annual international trade statistics detailed by commodities and partner countries.
Data URL: http://data.un.org/Explorer.aspx
Features of the table:
| Features | Name |
|---|---|
| 1 | Country |
| 2 | Year |
| 3 | Commodity_code |
| 4 | Commodity_description |
| 5 | Flow (import/export) |
| 6 | Trade Value in USD |
| 7 | Weight of commodity in Kg |
| 8 | Quantity measurement type |
| 9 | Quantity |
| 10 | Category |
The aim/acceptance criteria for the project was to successfully analyse international assets data to give meaningful insights and also test models like- Bayesian, Feed-forward neural network, ANOVA and statistical hypothesis testing.
python3
🐍
🌟 🌟 🌟
Pymc3, keras, numpy, pandas, scipy, matplotlib, seaborn, sklearn, plotly, squarify,
ANOVA, One T testing, Two T Testing, Principal Component Analysis, Neural Networks, Statistical Hypothesis testing, Visual Analysis
- Jyotsna Khatter
- Parth Gargava
This project is licensed under the MIT License - see the LICENSE.md file for details
Made with ❤️ at Northeastern University