I have an MS in Applied Math and work in machine learning (for the past 3 years). The only real advice I can give is I am not sure how much a BS will help (it certainly wont hurt) you with getting a job in machine learning. I ended up here but I spent my entire MS studying wavelets and signal processing. I did take a few courses in machine learning, but nothing super crazy.
I think if were in your position, I would spend six months taking free courses, writing some open source projects, making some cool visualizations with d3.js, and then leveraging that to try to get a job (or transition into a job). If I was hiring someone for a position in ML at my company, I would be more interested in seeing that type of stuff than asking questions like "How does an SVM work" or "Where did you get your BS in Math from?".
With that said, if you can afford the classes, have time to put in (a lot of time to really understand math), and you have always wanted to get a degree, go for it, and dont let anyone here stop you.
I think if I were only worried about a job I wouldn't go at all as I recently moved onto the Business Intelligence team at work because of how passionate and excited I am about the subject. I want more than I job, I want to deeply understand what I'm doing and excel at it. I'm going less for the short path (that I took in programming) and more for the most enriching.
I think if were in your position, I would spend six months taking free courses, writing some open source projects, making some cool visualizations with d3.js, and then leveraging that to try to get a job (or transition into a job). If I was hiring someone for a position in ML at my company, I would be more interested in seeing that type of stuff than asking questions like "How does an SVM work" or "Where did you get your BS in Math from?".
With that said, if you can afford the classes, have time to put in (a lot of time to really understand math), and you have always wanted to get a degree, go for it, and dont let anyone here stop you.