I don’t mind being asked technical questions, but I had an interview where I couldn’t figure out what the interviewer was going for. I think he started with, “let’s say you have a variable x, a vector, and pairs for (x,y). How can you predict y?”
I started by asking things such as, “is x all continuous or are there categorical variables?” Eventually the question got simplified down to a single x (scalar). I was explaining that if you plotted the x,y pairs you could draw plumb lines between your values and the line you’d predicted (least squares difference).
But no matter what I said, it seemed like he said, “no, just speak math to me.” I don’t know if he was looking for a particular term or some statistical theory. I felt like I was searching for a term that I couldn’t remember the name for, so I was describing things instead.
Residuals, maybe? Or assumptions you make when you do linear regression? I think he may have asked me a question that I simply did not know the answer to. But in the end, I couldn’t figure out his question. I haven’t even been able to go and learn or review whatever I was missing.
Another time, I was given a simple “write this function” problem. I knew the answer but the interviewer zoned in on the mistakes I made. It just dragged on and on, and it was a simple question that I was able to explain thoroughly and even say why my mistakes were wrong.
I think maybe I don’t know anything after all? I studied and practiced. Nobody ever really answered my question of how you know you’re qualified. Do I read all of SWEBOK? CLRS? Should I take a statistics class? Where do I find projects to work on that’ll help me “level up”?
Seriously, though, I still want to know the answer to that regression problem from my phone interview.