The most valuable thing you can get out of any book is a list of other books worth reading. Over time, you will find that certain books keep popping up from the "bibliography", and you should move those books to the top of the reading list. Other books will sink. Since the reading list is essentially a priority queue, you'll eventually find that some books sink so far down the queue that you may never read them again.
Start by picking a broad book to give you a general idea of the subject you are aiming for. Then choose a few specific books to master the aspects of the topic that interest you.
Reading the right book at the right time will have a better effect. A lot of reading can improve your reading ability. You will also improve your knowledge of the outside world through reading. For computer science, reading a lot of the latest information will be of great help to your future work and research. Reading will improve your horizon and enrich your spare time.
Even if you read a good book on programming every two months, which is about 35 pages a week, it won't take long for you to gain a deep understanding of our industry and differentiate yourself from everyone around you. -- Steve McConnell, Code Complete
Focusing on the classics also carries a risk: you devote too much energy to them, at the expense of more pragmatic knowledge and information that would improve your everyday skills. Make sure to mix classics with modern, more pragmatic books and articles on your reading list.
In practice, algorithmic problems do not appear at the beginning of a large project. Then, all of a sudden, they appear as subproblems when the programmer doesn't know how to proceed or when the current program is clearly inappropriate. -- Steven S.Skiena, Scenario Analysis of The Algorithm Design Manual