Projects Circle
The projects circle was developed to provide members with end to end data science project experience. In each cohort the participants are broken into teams. The teams support each other in working through all the ins and outs of taking a data science project from start to finish. Our project circles are a perfect way for you to hone your skills, meet new people in the industry, find mentorship and give back to the community.
Projects Circle Organizers
PAST PROJECT: SEATTLE TRAFFIC COLLISIONS DATA
This project used the Seattle Traffic Collisions data from 2004 to the present. Machine Learning as well as a Data Viz team were created to work with the data on the project. You can find all the details about the project at its Github repository.
Teams that finished the project presented their work on September 15, 2020. You can find the recording of the presentation here.
Past Project: Fraud detection competition
In April 2019, we launched our very first, mentored Kaggle competition! Senior Data and Research Scientists from Convoy gathered to mentor 4 different teams comprised of members of the Data Circles (formerly Sea-WiDS) meetup group. They worked on the TalkingData AdTracking Fraud Detection dataset to gain more experience doing predictive models on highly imbalanced data.
“I wanted to get more experience working on an analysis from start to finish. And I wanted to take advantage of the mentorship opportunity from Convoy. This project was challenging and frustrating, but overall I learned a lot.”
The teams met every Thursday for four weeks as a part of the Interview Prep meetup. Mentors were in touch with the teams via Slack and even met during the weekends. The teams trained a model and submitted their predictions to the Kaggle leaderboard, being scored on the top 25%. The final presentation of the results took place at the Convoy headquarters and if you are curious you can access the slides here.
There were many lessons learned during this challenge. Participants got to work with a real life problem, overcoming the size of the data, the imbalance, working with a time series problem all while learning about collaboration. Apart from the experience the participants now have one project on their portfolio that they can talk about during interviews.
Thank you to Convoy and our teams! We are looking forward for our next challenge!