Our AI hackathon brought together a diverse group of participants, who collaborated to develop a variety of impressive projects based on:
This event has now ended, but you can still register for upcoming events on lablab.ai. We look forward to seeing you at the next one!Checkout Upcoming Events →
Submissions from the teams participating in the Sustainable AI Hackathon (Swiss AI) event and making it to the end 👊
This is a Bitcoin Transaction Fraud Classification project Based on Elliptic Dataset provided by Elliptic Ltd. a London - based Blockchain analysis provider. Here is our Deployed Application link: https://share.streamlit.io/alihussainia/swissai/main This Project is built using Python based libraries. For data pre-processing we have used NumPy and Pandas, for data visualization we have used matplotlib and seaborn, for development we have used PyCaret and finally for web-development we have used Streamlit and StreamlitCloud. The Elliptic Dataset is a graph network of Bitcoin transactions with handcrafted features. All features are constructed using only publicly available information. Thank you for your Time and Consideration.
An app that lets you reduce waste by allowing the sharing of excess foodDEMO: https://share.streamlit.io/diegobiagini/saia_hackaton/main.py
The Next Convolution
One of the main source of greenhouse gases is heavy industry. Although industries are the driving forces of the economies, here we question how those industries and companies in them achieve to be fuel of machine called economy. In our concept we aimed to find relation between energy consumption and CO2 emission. Using this pattern we believe that an organization's environmental impact can be evaluated, which might be implicit. In the meantime, we are aware that our concept can be base for further studies, but cannot be definitive at this stage.
I decided to study the impact of earthquakes in my country (Chile) and how to mitigate it. For that I came up with an idea called "Earthquake Guardians". It implies that some people (suscribed to a specific Twitter account) will look at the messages sent of earthquake alarms and decide to confirm or deny such alarm. The account will send messages based on a deeplearning model trying to predict earthquakes by using the Grillo open database https://registry.opendata.aws/grillo-openeew/. Since earthquake prediction can be tricky, we count on these guardians to let us know if the earthquake predicted is real or not (and in the process, label the data for future training). Of course, earthquake prediction is not a new idea, but with the guardians and their help, it can create a better system with constant feedback. Also, in highly seismic areas (like Chile) the smaller earthquakes could also be detected and taken into account for future training of the model. I wanted to mix the idea of deeplearning prediction in python, with the Twitter solution like this one (https://blog.twitter.com/en_us/a/2015/usgs-twitter-data-earthquake-detection). Sadly, I didn't have enough time but to only do some EDA with the data.