Python 3, Scikit Learn, Pandas, SQL, Tableau
The goal of this project was to successfully build a prototype model and product that can be used as a tool for business and organizational research on deciding upon their viability and success in opening up shop.
What is Business Viability
A designation of determining if a business can successfully open shop with minimum barriers of entry.
For-Profit Viability – if a For-Profit business is able to open shop with minimum costs.
Non-Profit Viability – if a Non-Profit business is able to open shop based upon there being an unmet need within an area.
The Process was simple enough, clean the data and spend the first two days wrangling the data. Imputing was the crucial part, I did a simple means imputation for the missing values that were integers in the data. Testing and modeling dominated most of my time. Towards the end of this process was the visualization.
Understanding Viability Determination
The determination made was a binary one, a Y or an N for both non-profit viability and for-profit viability. In some counties you will notice that they aren’t viable for non-profits, but are perfectly fine for for-profit organizations. This has to do with certain variables as it pertains to child poverty rate and median household income. If income is above the national average and child poverty is below the national average then that county is not viable for non-profits. Non-profit viability as mentioned above is based upon need. For-profit viability is determined by the growth rate of fast food restaurants, full service restaurants, grocery stores, and convenience stores. If the determination is Y for non-profit but a N for for-profit in the same county might mean that the county is of need and business growth is not possible within that county. Such areas are classed as food deserts.
The original purpose was to identify food deserts, but upon wrangling the data and cleaning it, I found that I could develop a more versatile research tool.