DropModel

DropModel's landing page
User data is stored via an accounts system
Our step by step process will help you create your custom ML model
See your models train in real time!
Run predictions using your your personal AI model.

Inspiration 💡

Machine learning is a dynamic and rapidly evolving field within technology, characterized by its immense potential and countless opportunities. However, it can also appear daunting, given the overwhelming volume of information and choices available to those interested. This is where DropModel comes into play. The purpose for this project stemmed from the desire to demystify and democratize machine learning, making it accessible to anyone with an interest in learning - regardless of their experience or technical background.

What it does 🤔

DropModel simplifies the intricate process of training machine learning models into a user-friendly, three-step experience:

Upload your data: Users can effortlessly input their dataset and upload it to our platform, which will then convert the file. In this situation, data will be entered in the form of a CSV file.
Select your prediction variable: With just a few clicks, users specify the target variable for their model, defining what the model should predict. This makes it easy for the program to calculate the probabilities associated with each component of the model, and return the correct data back to you.
Sit back, relax, and enjoy: DropModel automates the labour-intensive task of model training, delivering efficient and highly accurate machine learning models as a result. Use this for any of your projects, or to learn more about machine learning and to feed your own curiosity. Machine learning is for everyone!

How we built it 🛠️

Creating DropModel required the use of several technologies, combining the strengths of TensorFlow, Python, React, Firebase, and FastAPI. TensorFlow served as the backbone for machine learning capabilities, Python facilitated the backend logic, React ensured a responsive and intuitive front-end interface, Firebase handled user authentication and data storage, and FastAPI enabled seamless communication between components. This stack was applicable to all of our strengths and experiences as a team and allowed us to create a great and complex project.

Challenges we ran into 🧩

Building DropModel was not without its challenges. We encountered the common issues of debugging and model training failures, which are inherent to the complexity of machine learning. Mathematical complications caused some errors within the program, but we were quickly able to resolve these in order to continue efficiently building the model training program. Additionally, the sleep-deprived environment of a hackathon posed its own set of challenges, highlighting the importance of managing time and maintaining team well-being during intensive development.

Accomplishments that we're proud of 🌟

Our journey with DropModel was one of continuous growth and achievement. The project seemed very ambitious initially, and some of us had concerns about whether we could even pull off a simple product in time. However, as we worked we became more aware of how we worked together and we were able to begin working at a faster and more efficient pace. In the end, we were able to add extra features and design, which we believe contributes a lot to the project as a whole. As a team, we are so proud of how we've done over the course of this hackathon, and we are happy to say that we have far surpassed our expectations.

What we learned 🧠

DropModel was like an intensive crash course in machine learning. Some of us had little to no experience with machine learning, so we dove headfirst into learning all we could. Each member of our team was able to learn on the fly and adapt quickly in order to use the technologies required for us to complete this project.

What's next for DropModel 🚀

As a team, we envision a future where DropModel becomes a household name among developers. To achieve this, we plan to open the gates and provide APIs, allowing developers to seamlessly integrate DropModel into their projects. Our mission to make machine learning accessible to all remains the same as we continue on the journey of refining and expanding this platform in order to help others pursue their interests in this fascinating field.