Reddit reviews Programming Collective Intelligence: Building Smart Web 2.0 Applications
We found 18 Reddit comments about Programming Collective Intelligence: Building Smart Web 2.0 Applications. Here are the top ones, ranked by their Reddit score.
O Reilly Media
Introduction to Algorithms is a behemoth text book. I prefer O'Reilly's Algorithms in a Nutshell and also Programming Collective Intelligence" for basic ML stuff.
I'd grab beautifulsoup + scikit-learn + pandas from continum.io (they're part of the standard anaconda download), launch Spyder and follow through this:
http://sebastianraschka.com/Articles/2014_naive_bayes_1.html
You can get a RAKE impl here too : https://github.com/aneesha/RAKE
Doing recommendations on the web like that is covered in an accessible way in "Programming Collective Intelligence"
I've posted this before but I'll repost it here:
Now in terms of the question that you ask in the title - this is what I recommend:
Job Interview Prep
Junior Software Engineer Reading List
Read This First
Fundementals
Understanding Professional Software Environments
Mentality
History
Mid Level Software Engineer Reading List
Read This First
Fundementals
Software Design
Software Engineering Skill Sets
Databases
User Experience
Mentality
History
Specialist Skills
In spite of the fact that many of these won't apply to your specific job I still recommend reading them for the insight, they'll give you into programming language and technology design.
In addition to BeautifulSoup there's also Scrapy if you want to do some crawling and screen scraping. http://doc.scrapy.org/en/latest/intro/overview.html
You might consider this book for a starter into data mining and machine learning. It uses Python for the code samples.
http://www.amazon.com/Programming-Collective-Intelligence-Building-Applications/dp/0596529325
I liked Machine Learning For Hackers, Programming Collective Intelligence and The Elements of Statistical Learning.
Seconding collaborative filtering. It's also a fairly simple algorithm to implement yourself as long as you're not using Wikipedia as a guide.
Collaborative filtering is like what Amazon uses to figure out what products to recommend to its users. It finds users that have similar purchasing habits to yourself and recommends items that they bought.
The first chapters of Programming Collective Intelligence describe how to implement Collaborative Filtering in Python in a really intuitive way, along with providing source code. Two hours in and you'll have a working service recommendation system. I'd definitely recommend that book to anyone looking to build what OP is interested in making.
If you want a quick non-textbook to get your feet wet, Oreilly's Programming Collective Intelligence isn't half bad.
Programming Collective Intelligence - but it uses python.
O'Reilly has published a number of practical machine learning books such as Programming Collective Intelligence: Building Smart Web 2.0 Applications and Natural Language Processing with Python that you might find good starting points.
Also machine learning, you profile sounds pretty good for machine learning. Do check out Andew Ng's videos, and this book. Machine learning is very much in demand right now, from AI, computational biology, finance, there's hardly any area where it isn't being used.
I swear by this book for an introduction to GAs and a ton of other cool ML/AI algorithms. No advanced math/probability knowledge necessary; it's focused on practical examples and intuitive explanations. It's an excellent foundation for further study.
Collective Intelligence
Sounds like you're running into O(n^2) or O(n^3) blowup. You didn't describe what algorithm you're using. Which probably means you don't know it, which means you don't know what the complexity is.
You need to make an index by item recommended. For speed, do it in C++ (e.g. a simple hash_map), but Python will be good to play with the algorithm.
Try posting 1M rows and I bet someone here (including I) could write something simple quite quickly.
Also try: http://www.amazon.com/Programming-Collective-Intelligence-Building-Applications/dp/0596529325
Although I don't believe they directly addressed algorithmic complexity. They presented some n^2 algorithms without really saying so.
FWIW , You might enjoy Programming Collective Intelligence if you liked this talk.
link to buy off author's website
I'm a general engineer myself, with a side interest in computer science. Szeliski's book is probably the big one in the computer vision field. Another you might be interested in is Computer Vision by Linda Shapiro.
You may also be interested in machine learning in general, for which I can give you two books:
I see you're interested in compilers. The Dragon book mainly focuses on parsing algorithms. I found learning about Forth implementations to be very instructive when learning about code generation. jonesforth is a good one if you understand x86 assembly.
I figure a business background and are looking to incorporate machine learning/AI into your pipeline. Programming Collective Intelligence: Building Smart Web 2.0 Applications is a must-read. Doesn't go too much into it but still gives you a good idea of the popular ML techniques and how they're being used by top companies.
If you want to learn the algorithms by programming them you have Programming Collective Intelligence that is really good. It really helped me to see the algorithms in work in order to deeply understand them.
Here are some links for the product in the above comment for different countries:
Amazon Smile Link: this book
|Country|Link|
|:-----------|:------------|
|UK|amazon.co.uk|
|Spain|amazon.es|
|France|amazon.fr|
|Germany|amazon.de|
|Japan|amazon.co.jp|
|Canada|amazon.ca|
|Italy|amazon.it|
|China|amazon.cn|
To help donate money to charity, please have a look at this thread.
This bot is currently in testing so let me know what you think by voting (or commenting). The thread for feature requests can be found here.