There has been few monthes that I haven't read PRML.
Few monthes ago, I tried to read the PRML book, I gained lots of insights of Bayes's view, however I spent lots of time. After reading the first four chapters I found it's really hard for me to move on, and through talking with others for few times, I realised that I may need to capture some frequently being used models or algorithms firstly, so I spent about two weeks to learn the Machine Learning's course: css229. During this period I learned more about deep learning, and took a reiew of LDA topic model. From these learnings I really learned lots of commonly used models, and while learning these models, sometimes I can only capture some easy parts, or while reading reference materials I may feel kinds of diffcult, what's more, I knowed what I need to learn and what I want to learn.
After these two days' communicating with others, I felt that perhaps I need to return to PRML and MLAPP to gain more insights about kinds of fields that I haven't understand well, like variational EM, RBMs etc, so I'm going to make a plan to read these books, especially for these parts. And after reviewing PRML just now, I felt really exciting, for the reason that, I know what each chapter mainly talking about. 以前看PRML像是在盲人摸象,现在终于能有目标的看了。