Our inspiration came from our current situation. We are at the core of our learning, and we have a deep interest for data science. What better way to learn the basics than start with a tool that will help us in the future! As college students coming from various parts of the world, we understand and have been in the situation where we are trying to find good housing that is in the proximity of our college, is safe, and affordable. We took up this challenge as a great way to put our experience to use and learn how to use data science to make a prototype of a great tool that can help various people.
What it does
For our specific project, we limited ourselves to picking cities within the United States and chose over a hundred cities to gather data about. For college students, we thought that the cost of living index and safety would be the two most important aspects to think about when finding a place to live. Using datasets ranging from the last couple of years, we created a map that plots all the cities, and when you hover your mouse over the points, you can see the data we have collected in small and readable tidbits! We also have a sliding bar in the top right hand corner of the project that allows you to filter the cities based on the cost of living index.
How I built it
We used python to sort our data by cities. We chose the most common cities that frequented our data and pulled only the data for those cities to output on our map. Once we condensed our csv files, we uploaded it to Tableau and created a user friendly interface that allows a person to look at the map and see an overview of the cities of their liking. We used the built in template to generate the map and data points, and added filters to allow the sorting of data.
Challenges I ran into
We faced a lot of challenges finding credible resources, sorting the data, and finding a good environment to set up a visual interface. We spent hours looking for data until we found two or three resources that we could depend on to get the information we needed, and once we got the data, we struggled a lot to understand how to use python to sort and merge our collected data. Most of us were new to python, so this proved to be the biggest challenge of all. Finally, we were debating whether to use the Google Colab environment to create a UI for our program or create a webpage from scratch. However, these option soon proved futile, and we had to ask for help. Once we got some guidance, we found Tableau to be a good resource to build or project on.
Accomplishments that I'm proud of
We are really proud of the user interface that we created, and the way we sorted our data. Given our minimal knowledge of data science, python and Tableau, what we managed to produce in the time frame that we had was quite impressive for us. We did not expect us to get this far or even create a presentable form of out project, and that itself is a great achievement. We also used the platform JIRA to split up tasks and get work done, and that allowed us to know what everyone was doing, where we were at in terms of the project, and what to get done. Our teamwork using all of these softwares and staying on task is another thing we are immensely part of.
What I learned
We learned a lot about basic python, machine learning, and Tableau. Although we did not end up using any machine learning techniques in our project, the workshops we attended gave a huge insight to a plethora of knowledge we did not know before. It was interesting to see how we could use our basic skills in mathematics and apply it in a way that predicts the trends of a given set of data. More related to our project, we learned how to merge data together and analyze what the values meant in terms of our project to increase the interpretability of our project. Tableau was a great platform to learn how we can take our computed data and make it more user friendly and accessible to other people across the globe in a way that can make a difference.
What's next for Find My Home
There is still so much room for growth for our project! We hope to refine our project to allow more data that we collected to be available on the map. We also would like to increase the amount of filters so that the user can see the results more visibly. Once we have these options available, we are planning to upgrade our model to allow the program to think for itself and predict what the city will look like years from today. Our longterm goal is to make the project efficient so that it can be scaled to an actual application available for users to use in the future.