Inspiration
As Data Science majors, the first-ever Datathon at UCSB represented a unique and exciting opportunity for us to be a part of history. We saw it as a chance not only to showcase our skills but also to further develop them in a competitive environment. The datathon being a chance to engage with a new tool SingleStore, was a perfect learning experience. Additionally working alongside friends added a layer of enjoyment to the learning process. Participating in UCSB’s first datathon allowed us to learn valuable in-demand skills, and although we only had a short period to learn SingleStore, this experience has motivated us to keep learning this tool and expanding our knowledge. As for our specific project, we decided to focus on analyzing the differences between good and bad wine, as we always heard the phrase “Aging Like Fine Wine” but never really knew what fine wine was, so using the dataset we found out what fine wine truly is.
What it does
Our project uses the power of data science to analyze the given dataset of Red Wine. Essentially, what it does is use various statistical techniques to go through and analyze the different physicochemical properties and sensory qualities of wine. Our goal was to find patterns and correlations to see what makes some wines good and others bad.
Using K-means clustering and decision tree classifiers, we categorized wines into different quality groups. We also created visualizations, like correlation matrices and distribution plots, to make the data more accessible and understandable. These visualizations help in identifying the key factors that most heavily influence wine quality, such as acidity levels and alcohol consent.
How we built it
With the use of SingleStore and their powerful software, we were able to create in-depth data analytics on Red Wine using the provided data set. We created a shared notebook and were able to create useful graphs using imported libraries such as pandas and matplot that helped us better understand the data and notice patterns within it. Because of SingleStores’ powerful capabilities, we were able to use both Python and SQL to get the necessary information to analyze what makes Red Wine so great. With extensive research on SingleStore and their tools and our background in Data Science, we were able to create meaningful information and make connections based on the visualizations and queries.
Challenges we ran into
Working with SingleStore presented its learning curve. As a new tool for my team, learning its functionalities was challenging. But through some YouTube tutorial videos and our previous knowledge of Data Science, we managed to navigate through and were able to leverage SingleStore’s useful tools to our advantage. The other main challenge we faced in our project was dealing with the dataset. The dataset was complex as it contained a diverse range of variables, something more difficult than we have previously been exposed to. Overcoming the challenge of analyzing the dataset taught us a lot.
Accomplishments that we're proud of
We are proud of our ability to learn a new tool. SingleStore was very useful and provided the tools we needed to accomplish our goal. Not only did we learn a whole lot about data science and a new tool but our project gave us some very useful and important data.
This project offers valuable insights into the world of wine quality. The information collected from this offers both consumers and sellers the best data to help guide decision-making. For consumers, it is about making informed decisions when buying new wines to make sure they don’t waste money on bad quality wine. For the sellers, our data provides helpful information about which wines to stock that meet the quality the customer may expect.
What we learned
The biggest things we learned were how to use SingleStore and the technical skills needed for this project. We learned how to effectively manage and analyze a complex dataset. And learned different types of data visualization to help make the data more presentation and easier to understand. Besides the technical skills, this project taught us to use the skills of teamwork and communication. We learned how to divide work among our team so every teammate would be doing something productive and how to combat errors when multiple different people work on the same code.
What's Next for Age like Fine Wine
This Datathon has been a massive learning experience and taught us many useful skills as well as expanded those that we had previously. Given more time, our team has said the next thing we would have to do with this project would be to make the data more user-interactive, so the user could engage with the data.
As for what is next for us as students, this experience has helped us understand how working on a team is for programming. We will continue to learn more about data science in our coursework and apply this newly learned knowledge to our future classes and careers. We hope to be able to participate in a datathon like this one again to keep learning data science to keep fueling our passion.
Log in or sign up for Devpost to join the conversation.