Every Tuesday in April: Free Intro to Python Training Series for Business Professionals Register Now

Made at Metis: Classifying Car Images; Navigating Media Bubbles with Data Science

By Emily Wilson • December 04, 2019

This post features two projects from recent graduates of our data science bootcamp. Take a look at what's possible to create in just 12 weeks.

Classifying Car Images Using Features Extracted from Pretrained Neural Networks
Bhanu Yerra, Metis Seattle Bootcamp Graduate

For his 3rd (of 5) bootcamp projects, Bhanu Yerra explored how a car image classification system could address the following business issues: 

  1. Ground-truthing of posted used car images on peer-to-peer sales platforms - are these images actually the cars they specify? Do multiple exterior pics represent the same car?
  2. Organizing web-page displays based on the user uploaded images
  3. Recommending alternative cars available in the inventory that have similar looks and price

In addition, he wrote in a blog post about the project, "a simple classification system of cars can help in identifying the fine-grained features important for 3D object detection for self-driving cars." Read the full blog post here to learn about his approach and results. 

_____

Navigating Media Bubbles with Data
Aisulu Omar, Metis Seattle Bootcamp Graduate

It's no secret that we live in a world – and broken down even more, in the U.S., we live in a society – with significantly sized media bubbles, from which many get their news. While a student at Metis, Aisulu Omar thought about this a lot and decided to focus her final project on the topic. From her time traveling to and living in various places, she realized how much these bubbles contribute to people's sense of reality.

"The more I read news coverage in places with different media bubbles, the more I became sensitive to biases in media by reading news coverage. This cultural aspect creates a disconnection between people," she writes in a blog post about the project. "I think for people to connect and have intelligent conversations, it would help to be aware of different media bubble’s narratives on any given topic. This has prompted me to create a news recommendation platform that provides narratives from different publications."

She began by first setting out to prove that the media has these bubbling biases, and then she proceeded to build the recommendation platform. Read the post in full to get all the details.

_____

See more examples of Metis student projects here


Similar Posts

data science
Our Top 10 Most-Read Blog Posts of 2019

By Emily Wilson • December 20, 2019

Throughout the year, we post blog content including alumni stories, data science insights from our Sr. Data Scientists, guest posts, and much more. Check out our top 10 most-read posts of 2019. We hope you enjoy them again or for the first time, and we look forward to producing much more data science content in 2020.

data science
5 Common Questions from Bootcamp Applicants

By Emily Wilson • February 17, 2020

While there are a number of ways to enter the data science field, we believe an immersive bootcamp is the most efficient and effective path with the best overall return on investment. But understandably, many people have questions about how a bootcamp will help them navigate the process of landing a job in the field. In this post, we share answers to 5 common questions that we receive from potential applicants.

data science
Made at Metis: Video Search Engine Using Salsa Dance Videos & Predicting Strategy in StarCraft

By Metis • January 13, 2020

This post features two projects from recent graduates of our data science bootcamp. Take a look at what's possible to create in just 12 weeks.