SC1015 Lab Group SC14 Mini-Project focusing on "World Happiness" datasets from the years 2015-2021.
Our aim is to help countries increase the well-being of their citizens. One way they can do this is by making use of "The World Happiness Report", a publication of the United Nations Sustainable Development Solutions Network. It contains rankings of national happiness, based on respondents' ratings of their own lives, which also correlates with various quality of life factors.
In order to create good policies in the present, how can countries predict and use their world happiness statistics in the future?
1.Download the required files and folders from this repo: notebooks, uncleaned_data, plotly and requirements.txt.
2.Run the command pip install -r requirements.txt in your jupyter environment.
3.Explore eda.ipynb. This notebook includes:
- Exploratory Data Analysis
- Machine Learning (Part 1)
4.Explore cleaning.ipynb. This notebook includes:
- Data preprocessing
- Data cleaning
- Data exporting
5.Explore machine_learning.ipynb. This notebook includes:
- Machine Learning (Part 2)
- Recommendations
Overall, we have learnt a lot from this project. We discovered and utilised many new concepts that were not taught to us in labs. We managed to conclude that Autoregression(AR) on Score is the best model for forecasting. However, we also realised that we may have used the wrong approach by trying to utilise time series forecasting, which is not as effective when used on yearly data. This experience has been fruitful, and we will remember these new concepts for other data science projects going forward.
- ipython 8.2.0
- matplotlib 3.4.3
- numpy 1.21.2
- pandas 1.3.3
- plotly 5.7.0
- scikit_learn 1.0.2
- scipy 1.8.0
- seaborn 0.11.2
- statsmodels 0.13.2
- Linear Regression
- Random Forest Regression
- Gradient Boosting Regression
- Ridge Regression
- Huber Regression
- Autoregression
- Autoregressive Integrated Moving Average on Score (ARIMA)
- Vector Autoregression (VAR)
- Adapting and using concepts taught in lab
- Correlation Matrix
- Pair Plot
- Diagnostic Plot
- Data cleaning concepts
- Creating 136 datasets after merging 7 datasets
- Converting dataset from non-time series to time series
- Representing numeric data as a geographic map in plotly
- New machine learning models and concepts
- K-Fold Cross Validation
- Random Forest Regression
- Gradient Boosting Regression
- Ridge Regression
- Huber Regression
- Time Series Forecasting
- Autoregression
- Autoregressive Integrated Moving Average on Score (ARIMA)
- Vector Autoregression (VAR)
Andrew Oak Soe Khant @xaynezz
- Code Refactoring, Exploratory Data Analysis, Data Cleaning, Regression Models, Time Series Models, Slides, Video Editing
Oza Tirth Tusharbhai @Tirth3033
- Exploratory Data Analysis, Slides, Presentation
Sisodia Anushka @anushkasisodia
- Exploratory Data Analysis, Slides, Presentation
Thanks to Dr Sourav Sen Gupta and TA Song Nan for their teaching and guidance.
- https://www.kaggle.com/datasets/mathurinache/world-happiness-report
- https://plotly.com/python/choropleth-maps/
- https://machinelearningmastery.com/k-fold-cross-validation/
- https://www.analyticsvidhya.com/blog/2016/02/complete-guide-parameter-tuning-gradient-boosting-gbm-python/
- https://www.machinelearningplus.com/time-series/arima-model-time-series-forecasting-python/
- https://www.machinelearningplus.com/time-series/vector-autoregression-examples-python/
- https://towardsdatascience.com/regression-in-the-face-of-messy-outliers-try-huber-regressor-3a54ddc12516
- https://towardsdatascience.com/ridge-and-lasso-regression-a-complete-guide-with-python-scikit-learn-e20e34bcbf0b
- https://towardsdatascience.com/machine-learning-basics-random-forest-regression-be3e1e3bb91a