I’ve gotten very into brains lately, and I’ve been inspired to talk about them. I find giving talks a good way to be inspired to deep-dive on a selected topic and organize my thoughts such that I can explain them to others. I’ve given a lightning at the last two Learn Data Science Lightning Talk nights. You can find the videos below. I think the second talk is definitely better than the first, and I will continue to improve with practice.
A link to the second video, since WordPress won’t let me embed multiple YouTube videos: https://www.youtube.com/watch?v=0d-eM2FToZg
The competition was put forth by The Nature Conservancy. The task was to automatically detect and identify fish from still images taken from cameras mounted on fishing vessels. This would help in the enormous effort of monitoring fishing activity in sensitive areas.
I competed on a team (Fishy RSeaNN) for this competition. We placed 84th out of 2293 teams. This put us (barely) in the top 4% of the competition, but we still performed worse than the sample submission for the competition.
This presentation discusses the competition and its challenges, some winning strategies used by the top-10 winners of the competition and some state of the art techniques in object detection and classification within images. If you want to follow along with the slides, you can find them here.
(Apologies for the number of “umms” in this video. This is a great example of why you should record yourself giving presentations so that you can hear yourself and your presentation style before you present in front of other people. Regardless, if you can get past the filler, the content is pretty good. Enjoy!)
Well, I have submitted the official form, so I am officially dropping out of the Big Data Master’s program at SFU. It seems kind of ironic that the girl who was featured in the promotional video for this year’s cohort is now leaving the program halfway through. I wanted to write this post to explain why.
I am not leaving because I’m not doing well.
I had a sweet co-op job lined up at IBM (that I found myself. Yay networking!) and an A- average from last term’s courses (including an A+ in Machine Learning!). So let’s just get that out of the way.
I have milked the program for all it’s worth to me.
This master’s program is front-loaded with courses; 6 out of the 7 required courses are completed in the first 8 months of the program. The latter half of the program consists of only a co-op term and one additional course. For me, the one course that I will leave incomplete is a course in algorithms, the equivalent of which I have already taken during my undergrad. The co-op is not extremely relevant since I have extensive past co-op and full-time work experience. So really, there is nothing more that this program can give me that I cannot get myself, other than a shiny piece of paper.
I already have proven co-op experience in data science.
During my undergrad, I completed 5 co-op terms (a total of 20 months of experience). My final co-op job was an 8-month stint at the Ontario Institute for Cancer Research, applying machine learning techniques to somatic mutation data in order to subtype cancer patients. It was this position that first got me interested in data science. I spent 8 months learning about different algorithms and techniques, reading and applying ideas from academic papers and writing code that worked on a distributed system.
OMG this program is expensive.
The fee structure for this program in a strange way compared to what I’m used to. During my undergrad, in addition to our tuition we paid a set co-op fee (~$600) during our study terms when we were taking classes and applying for jobs. In this program, we pay a set fee per term (~$7000) for 4 terms. This includes our co-op term. To me, this is equivalent to paying a ridiculous sum for the privilege of working. This tuition fee is also set regardless of the number of courses you’re taking. So in the last term, where you’re only required to take one course, you are paying $7000 for this privilege. I have already paid 2 semesters’ worth of tuition, and while I am grateful for the knowledge and experience I have gained from these semesters, I already feel like I have overpaid.
I’m really happy with my decision to withdraw from this program. I’m excited to find a full-time position and start doing real-world work in data science.
I recently gave a talk on building neural networks using Python’s Keras library at the Vancouver PyLadies Meetup Talk Night. You can check out the slides here.
On March 4th, I participated in Open Data BC’s Open Data Day hackathon. My team (Access360) built a map interface to display points of interest for those with accessibility issues. We mined open datasets from the City of Vancouver to find information on accessible parking spaces and accessible public washrooms (did you know there are only 16 such washrooms in the whole of Vancouver?!). We used API’s from Yelp and TransLink to gather data on accessible venues and transit stops in the City of Vancouver. We ended up winning first place in the Wildcard category (with a $500 prize) as well as the Protohack Challenge, winning us a team pass to Protohack. I had an awesome day learning how to process KML files and teaching my team members Python. You can find our project on our Github page.
