View Additional Projects

Air quality measures the amount of pollutions utilizing a number of indicators including the amount of particles (PM2.5 and PM10), carbon monoxide, ground-level ozone, sulfur and nitrogen dioxide. The rollup of these indicators are summarized in the AQI or Air Quality Index, The AQI is a measure from 0 to 500 that lists the severity of the issue. with +301 being hazardous and -50 being good.

However, overall our focus will be observing of what are the most common raspatory illnesses occurring within India, due to the poor air quality. And aiding in the decision of where the government should build new healthcare facilities with the assistance of private contractors.

These contractors want to look at a variety of factors - including where the AQI is increasing the fastest, but also where the population is growing the fastest. They want to ensure a good amount of traffic coming through the healthcare facilities to make the building of them worthwhile.

See full report [HERE](https://drive.google.com/file/d/1pGCPD76i8MGmrU1dUx7NvGoe28n6OTHZ/view?usp=share_link).

![](https://preview--energetic-robin-8249d.stackbit.dev/images/Screenshot%20\(100\).png)

![](/images/Screenshot%20\(101\)-462cbe25.png)



Air Quality in India and It's Impact in Healthcare

This project looks at the amount of the loan, their current mortgage amount, their assessed value, the reason for their loan, what category their job falls under, YOJ is years at their job (the longer the better), DEROG is number of derogatory reports (the lower the better), DELINQ is the number of Number of Delinquent Credits (the lower the better), CLAGE is age of oldest credit (older the better), NINQ is the number of credit queries (the lower the better), CLNO is the open number of credit lines (number can be debatable on if they pay on time or not), and DEBTINC is debt to income ratio.

I started my analysis by manually breaking this data down by Training (60%), Validation (20%), and Test data (20%) within a validation column. There is some randomness in the Neural Networks and Random Forests and Boosted Trees on where they start analyzing the data from so running this multiple times might not yield the same exact result.

The best part of employing the boosted trees modeling is that it learns from its mistakes and then rebuilds the model over and over putting more weight into these errors until it can no longer improve. Once it has reached it peek performance, it takes an average of the results. By using this average, it confirms that the predictions are stable.  

For the full report click[ here](https://drive.google.com/file/d/1VCyfdXwbDuuX_JGX6KJ2yEDju2k9DWR5/view?usp=share_link).

![](/images/Modeling-0f40b046.png)

![](/images/Modeling2.png)

Overall, the data shouldn’t be surprising for those who have applied for any type of credit line before. You know your credit score considers your length of history, the amount of utilization, any missed payments. I do like seeing the “bigger picture” of how those with larger mortgages and those who are taking out larger loan amounts seem to also be taking it out for home improvement reasons. This will just add more value to the home and lower your debt-to-income ratio which is another variable that does have an impact. Allowing those who are in good financial standing to improve their standing even more.

The lessons learned – pay off what debt you have, ones that are high interest or large dollar amounts, keep making payments on time. That way, when eventually one does buy a home and want to increase their equity and thus net worth it will really payoff.


Forecasting Model on Home Loans

I have seen many articles in the past year of how fertility rates have been dropping so significantly countries are paying people to give birth. It sparked an interest in how education and economies have shifted this change.

As the average years of schooling increases, the fertility rate drops. This might be because there is less time to focus on raising children when someone is still working through degrees. 

*   Which country has seen the most decrease in fertility rates since 1950?
    Singapore had an 80% decrease in Fertility Rates since 1950. With 6.4 average in 1960 dropping to 1.27 70 years later. This is also now one of the richest countries in the world. 



*   Are there any countries that do not follow the overall trend? 
    For all the countries we have data, globally all countries have increased their education level and fertility rates decreased



*   Can we predict that becoming more educated have an impact on future population growth?
    With education comes better knowledge and accessibility to birth control. I think eventually fertility rates will hit a bottom baseline but believe education has become more accessible for so many people that this will continue to grow. 

For the full report click [here](https://drive.google.com/file/d/11ESNJ97JL2JxFz2Dy0b4CfVLgYzJx1iM/view?usp=share_link).

![](/images/Edvsbrith-25db372c.png)![](/images/Edvsbrith2.png)

Relation Between Fertility Rates and Education Levels

Comparing home sales and rental prices in NC over a period of ~30 years we could see which areas recovered quickly after the 2008 Great Recession within North Carolina, and also to other major metro areas in the country. It also takes a deep dive into rental markets around NC.

For the full report click [here](https://drive.google.com/file/d/1OilDwFcEzZ6C11nvVYRzlE-kVoKFNA6K/view?usp=share_link).

Example of the median home prices in the 4 major metro areas within North Carolina:

![](/images/Screenshot%20\(105\).png)



Projects

Air Quality in India and It's Impact in Healthcare

Forecasting Model on Home Loans

Relation Between Fertility Rates and Education Levels

Home Value Index