The cold start problem: how to build your machine learning portfolio

I said in that post that one thing you should do is build a portfolio of your personal machine learning projects..Ron wanted to work at Company X so badly, in fact, that he built a personal project that was 100% dedicated to getting him an interview there.We don’t usually recommend going all-in on one company like this..But — like I said — Ron really wanted to work at Company X.So what did Ron build?The red bounding boxes indicate missing items.Ron started by duct taping his phone to a grocery cart..He did this 10–12 times at different grocery stores.Once he got home, Ron started to build a machine learning model..Finally, he built his loss-of-consciousness classifier.At the same time as he was doing all these things, Alex was showing snapshots his project to hiring managers at networking events..Every time he took out his project and showed it off (on his phone), they asked him how he did it, about the pipeline he built, and how he collected his data. But they never quite got around to asking about his model’s accuracy — which was under 50%.Alex planned to improve his accuracy, of course, but he was hired before he got the chance. It turned out that the visual impact of his project, and his relentless resourcefulness in data gathering, mattered much more to companies than how good his model actually was.Did I mention Alex is a history major with a minor in Russian studies?What they have in commonWhat made Ron and Alex so successful? Here are four big things they did right:Ron and Alex didn’t spend much effort on modelling. I know this sounds strange, but for many use cases nowadays modelling is a solved problem. In a real job, unless you’re doing state of the art AI research, you’ll be spending 80–90% of your time cleaning your data anyway. Why would your personal project be different?Ron and Alex gathered their own data. Because of this, they ended up with data that was messier than what you’d find in on Kaggle or the UCI data repository. But working with messy data taught them to deal with messy data. It also forced them to understand their data better than if they’d downloaded it from an academic server.Ron and Alex built visual things. An interview isn’t about your skills being objectively assessed by an all-knowing judge. An interview is about selling yourself to another human being. Human beings are visual creatures. So if you pull out your phone and show the interviewer what you built, it’s worth making sure that what you’ve built looks interesting.What Ron and Alex did seems insane. And it was insane. Normal people don’t duct tape their phones to shopping carts. Normal people don’t spend their days cropping pilots out of YouTube videos. You know who does that? People who will do whatever it takes get their work done. And companies really, really want to hire those people.What Ron and Alex did might might seem like too much work, but really, it isn’t much more than you’d be expected to do in a real job. And that’s the whole point: when you don’t have work experience doing X, hiring managers will look for things you’ve done that simulate work experience doing X.Fortunately you only need to do build a project at this level once or twice — Ron and Alex’s projects got reused over and over for all their interviews.So if I had to summarize the secret to a great ML project in one sentence, it would be: Build a project with an interesting dataset that took obvious effort to collect, and make it as visually impactful as possible.And if you have a project idea and you aren’t sure if it’s good — ask me on Twitter! My handle is @neutronsNeurons, and my DMs are open :)***************************************************************[1] In case you’re wondering why this is important, it’s because hiring managers try to assess you by looking at your track record.. More details

Leave a Reply