In my last few years, it's hard to tell what I've learned. Just a bunch of facts, but nothing amazing conceptually. From now on, I plan on reflecting upon/ remembering what I learn each day; what I learn each week; each quarter; each year. The key lies with actually learning something, conceptually speaking -- not just facts, although those are good too.
Today I learned a bit about machine learning:
- Supervised Learning -- train an algorithm with data, which consists of "inputs" and "outputs"
- Regression analysis used with continuous variables
- Classification used with discrete variables
- Learning Theory -- the theory of machine learning
- Unsupervised learning -- using an algorithm to identify "groups" of variables, without prior training or criteria
- Reinforcement learning -- identifying "good" and "bad" behavior, allowing the machine to adapt
No comments:
Post a Comment