yiranjing / health_and_happiness_study

Ananlysis possible relationship among Happiness, Mental Health, Body Health, Education, Pollution, GDP, Employment etc.

Health and Happiness Study

Data1002 group assignment!

Author:

• Eileen Wang
• Yiran Jing

Report

Report

Introduction

Pursuing happiness has always been a fundamental part of human nature, with some arguing that it is the reason why we strive for success, maintain relationships and do more than just exist. Philosophically and semantically, one can easily define happiness and describe what constitutes this feeling. However, given that happiness is an intangible feeling that is subjective depending on the individual, happiness is often difficult to compare and measure, especially between people among different cultures. Nevertheless, past studies and literature have attempted to quantify happiness using a combination of subjective measures such as responses from surveys and objective numerical measures like life expectancy and financial measures. Such studies have given rise to the term ‘happiness economics’, an academic field which grew extensively during the late 20th century. ‘Happiness economics’ or the ‘economics of happiness’ is the theoretical and quantitative study of the relationship between happiness, well-being, or life satisfaction with factors derived from sociological and economic indicators. In our research, we analyse a dataset containing happiness scores for each country with corresponding subjective and objective measures in an attempt to answer the following questions:

1. On average, is there an increasing or decreasing trend in happiness scores overtime for specific countries?
2. Which countries and which specific regions tend to have higher happiness scores?
3. What determinants lead to higher happiness scores for a country?
4. Do nations that report higher/lower levels of happiness demonstrate consistent patterns in factors that influence happiness?
5. What are some other interesting relationships/insights found in our dataset (that is not necessarily related to happiness scores)?

Prediction

1. Varibale selection
2. Linear regression
3. Regression based on k-nearest neighbours
4. Regression Trees
5. Model Comparison

health_and_happiness_study's Issues

Nine plots for pairwise Relationships with Happiness Score

I change the variables for ‘Pairwise Relationships with Happiness Score’ part, the current 9 variables are used in the model part. I select them since they looks have clear (linear) relationship with happiness score, however, KNN and decision tree still better than linear/ridge regression, which means that nonlinear pattern exists. Please justify them, as the variable selection part? Then I can just refer directly in part 3?

@EileenWang1996

havenot merge Suicide dataset

Not sure how to merge it

Lect content for marking criteria

These are Alan talked during this Monday lecture for stage 2: Please reference if you need

1. The report must itself to show criteria have been met
2. Make sure meeting “pass” criteria first. (If fail meet pass level, cannot get >50%)
3. Structure report for each section for easy reading
4. Part two should include all analysis apart from predictive models based on the actual data
5. Part one needs charts/tables with summarized actual data, can speak the predictive model if we get high accuracy. The audience is general ppl, who interested in relationship, pattern etc. from data.
6. Report two is how to build table/plot, the audience are IT person.
7. If draw only one-year data(for examle we select 2015 only), probably not four-way relationship
   @EileenWang1996