At its core, deep learning tries to mimic the process of learning in the brain. But the actual mapping happens with something called a neural network.
Last month, BrainStation hosted Demo Day for the graduating classes of the Spring 2019 Diploma programs. This graduate project showcase was one of the best yet, with hundreds of attendees and Hiring Partners joining us at our Toronto and Vancouver campuses.
We’ve rounded up some of the top projects from our Data Science, User Experience Design, and Web Development Diploma Program graduates:
Grover is a smart, personal plant assistant designed to help busy plant novices preview, select, and maintain houseplants. This app makes it easier to choose plants that will thrive in your space and gives you a better understanding of their behavior and needs.
Iskandar created theCook, an app built to engage a community passionate for healthy eating. Use the app to discover what world-class athletes and fitness gurus eat to reach their peak performance. See Iskandar’s portfolio to see more.
Calling is an app that works to bridge the gap between volunteers and organizations by helping volunteers connect with and apply to volunteer organizations. You can learn more about Zack’s projects on his website.
Kindred aims to bring together solo travelers to do two things: connect and explore. Travelers can discover a city’s best kept secrets through adventures designed around their interests, all while building meaningful connections with like-minded travelers. See Julie’s full profile.
Andrew uses reinforcement learning to train an AI to play PONG. Before training, the AI acts randomly, storing all the results of its actions in memory. After seven hours of training using a neural network, it is able to match the other AI PONG play for points. See Andrew’s portfolio.
Krypto simplifies the process for trading bitcoin and other digital currencies with an intuitive and seamless trading platform. See Michal’s design process on his portfolio.
Link was created to help foster strong connections between volunteers and organizations. Through strategic filtering, this app focuses on the volunteer’s passions and allows them to find like-minded companies.
Purchase Behavior Predictor
Lauren Taylor – Data Science Diploma Program
Lauren created a model that predicts consumer behavior on an eCommerce website based on a customer’s relationship with online channels (such as display ads, social, and organic search).
NBA Draft Picker
Eman built a classification model to predict the rank of top NBA prospects, using 10 years of NCAA player’s advanced stats, historical data, and draft results. Eman’s model allowed her to predict the results of the draft’s first two rounds with 82 percent accuracy.
Web Development Graduate, Joshua T., created Warcraft Auctions. The platform is a source for people who play World of Warcraft to view the game’s auction house details. Users can compare price and quantity history of auction house items to flip them and earn more gold. Want to see more of Josh’s work? See Joshua’s full portfolio or Github.
RescU helps people suffering from depression by allowing users to book training sessions with rescue shelter dogs so they can experience the benefits of animal therapy, group therapy, and a greater sense of life purpose. See more of Ivan’s work.
TL;DR Automatic Text Summarizer
Douglas Guglielmelli – Data Science Diploma Program
This automatic text summarization tool extracts the most relevant information from articles. Each sentence is represented by a node and the links represent how similar they are. Then, a ranking algorithm identifies which sentences are most relevant. See his portfolio.
Turn your photos into a graphic work of art with PhotoSprite. This web application converts your images into pixel art that you can then save to a gallery. To see more of Luke’s work, check out his Github Portfolio.
BCW Trailers: Instant Price Calculator
Bart extracted 11 years of market demand information from BCW’s sales quotes in order to provide more insight into their customers. Then, he trained a price calculation model to give an instant provisional quote which can be accessed by a web app. See the details of Bart’s project.
TexDio can take images with text and generate an editable text file or audio file in an array of languages. TexDio is beneficial for people with communication and learning challenges. It is also useful for educational purposes, such as learning a new language or improving listening and pronunciation skills.
Using historical sales data for 45 Walmart stores, Nimmy built a sales forecast model to accurately predict the sales of each department within each location taking holidays and promotions into consideration. The most accurate model was developed with 98% accuracy. See more of her work.