On Tuesday, October 18th, WISE (Women in Science and Engineering) Acadia presented a collection of poster presentations bringing together a variety of researchers from across campus. Each poster represented a unique way to look at how technology can be applied to different fields. In addition, the event included a talk by Dr. Anne Condon titled, “It All Computes: Celebrating Diversity in Computing”. The event took place in Fountain Commons, and brought together women from all scientific disciplines to address women in Computer Science. Each poster brought a good topic to the conversation, and the event overall encouraged female scientists to continue pursuing their degree of choice.
Some of the projects featured in the event include:
Girls and Gaming: Influencing Young Women to Join Computer Science (Hayly Thackeray)
The purpose of this thesis was to create a curriculum for girls in Grades 7 and 8 that teaches them the basics of Computer Science through creativity in video game design. The creative and fun game aspect will spark an interest in this age group, influencing their decision on participating in the Computer Science field. The Girls Get WISE Techxpedition was held on August 18th, 2016 with the help of WISE Acadia, WISE Atlantic, the Anita Borg Institute, Open Acadia, and the Jodrey School of Computer Science. The girls learned to create their own video game in Scratch and try out some Lego robotics. 25% of the girls that attended the program said they wouldn’t have considered becoming a Computer Scientist before the event, but are now. In addition, 75% of the girls said they would continue to program after the event, 20% said they weren’t sure if they would continue to program, and 92% of the girls said they enjoyed the event overall.
New Radial (Jon Saklofske, Marc Muschler, Ian Brunton and the INKE Research Team)
New Radial is a data visualization environment that offers an alternative digital workspace for visualizing images and text. It encourages users to contribute to ongoing academic dialogue via connections and operates as a middle ground between large datasets and a close reading environment. Radials are circular arrangements of related nodes that play a central role in a user’s exploration, collection, curation and connection of database objects for humanities-based scholarship. They allow users to work within a middle ground between large datasets and a close reading environment, providing an alternative scale of interpretation within a digitized framework.
Monitoring Mammal Movement Through Terrestrial and Aquatic Passageways in Antigonish, Nova Scotia Using Remote Camera Sensing (Stephanie White, Randy Milton, Trevor Avery)
Nova Scotia is gradually expanding its major 100-series highways in response to safety concerns. Highway twinning along a stretch of the Trans-Canada Highway intersects with an ecologically sensitive watershed, disconnecting wildlife movement in the area. To mitigate highway impact, wildlife terrestrial and aquatic passageways were installed along with 4 km of fencing and four one-way control gates. Passageways are >50 m long, and the terrestrial passageway has a central atrium. These passageways are the first to be studied in Nova Scotia, as well as the only to integrate both fencing and an atrium into the design. The highway is set to open late fall 2016, providing an opportunity to study wildlife movements before construction, during construction, and after installation. The scope of this 4-year project covers monitoring the movement of mammals through the passageways using remote camera sensing with a focus on medium-sized mammals (i.e. smaller than deer). Quantifying activity through images will be used to identify drivers of wildlife movement within the passageways. Factors under consideration include fencing, light intensity, moon phase, temperature and weather conditions. Select factors are manipulated to identify which may have a greater effect, including altering light levels by covering the atrium for extended periods connected to moon phases. Findings will be used as a baseline for future passageway design in Nova Scotia.
Escaping Local Minima with Symbols (Ahmed Galila)
This concept was inspired by how the brain processes sensory input. Inputs are propagated through the network one layer at a time. Each layer learns a certain representation of the input data. Each representation is more abstract than the layer below. This allows deep architectures to provide better generalization. The goal behind training deep architectures is to reduce the error in the model’s output as much as feasibly possible. Deep architectures have complex error functions with many local minima. Deep models tend to be trapped in these local minima.
In addition, the event included topics such as: Rethinking Time Course Data: Growth Curve Analysis in Spoken Word Recognition Research (Alexander Sproul, Randy Lyn Newman), Data Mining and Machine Learning at the Acadia Institute for Data Analytics (Danny Silver), and Unsupervised Multi-modal Learning (Mohammed Shameer Iqbal)
“I really liked the poster session because it gave people the chance to talk to people and ask questions, instead of going to a lecture.” – Acadia Computer Science major
The poster session was followed by a talk given by Dr. Anne Condon (a professor of Computer Science at U British Colombia). Her enlightening portion of the event discussed the difference between male and female enrollment in the area of Computer Science, and suggested some ways to combat this problem. Her discussion pulled together the importance of having programs such as WISE, and made the event even more exciting.
“If you get the chance, she’s [Dr. Condon] really nice to talk to, and she’s really smart. She really loves what she does, which I think is amazing because I’m struggling myself to figure out what I actually want to do with my Computer Science Degree… She has found her niche and she loves it. I went to her talk earlier, and it’s incredible how much she loves what she does. She’s really amazing, and I like that a lot.” – Hayly Thackeray