2. J. Leskovec, A. Rajaraman, J. Ullman: Mining of Massive Datasets, http://www.mmds.org 2

3. 3 Data contains value and knowledge

4.  But to extract the knowledge data needs to be  Stored  Managed  And ANALYZED  this class Data Mining ≈ Big Data ≈ Predictive Analytics ≈ Data Science 4

5. 5

6.  Given lots of data  Discover patterns and models that are:  Valid: hold on new data with some certainty  Useful: should be possible to act on the item  Unexpected: non-obvious to the system  Understandable: humans should be able to interpret the pattern 6

7.  Descriptive methods  Find human-interpretable patterns that describe the data  Example: Clustering  Predictive methods  Use some variables to predict unknown or future values of other variables  Example: Recommender systems J. Leskovec, A. Rajaraman, J. Ullman: Mining of Massive Datasets, http://www.mmds.org 7

8.  A risk with “Data mining” is that an analyst can “discover” patterns that are meaningless  Statisticians call it Bonferroni’s principle:  Roughly, if you look in more places for interesting patterns than your amount of data will support, you are bound to find garbage 8

9. Example:  We want to find (unrelated) people who at least twice have stayed at the same hotel on the same day  10 9 people being tracked  1,000 days  Each person stays in a hotel 1% of time (1 day out of 100)  Hotels hold 100 people (so 10 5 hotels)  If everyone behaves randomly (i.e., no terrorists) will the data mining detect anything suspicious? 9

10.  1 billion people  Everybody goes to hotel sometime in 100 days  Probability of 0.01 to visit hotel on a given day  Probability of 2 people choose the same day = 0.0001  Hotels hold 100 people  100,000 hotels  hold 1% of people  Probability that two will choose same hotel same day  0.0001 * 1e-5 = 1e-9  Choose same but different hotel on two different days = 1e-18  Approximate n choose 2 with n 2 /2  Number of pairs of people is 10 18 /2 = 5e17  Number of pairs of days = 10 6 /2 = 5e5  Probability of any one pair of people on a pair of days  5e17 * 5e5 * 1e-18 = 250,000  Expected number of “suspicious” pairs of people:  250,000  … too many combinations to check – we need to have some additional evidence to find “suspicious” pairs of people in some more efficient way 10

11.  Data mining overlaps with:  Databases: Large-scale data, simple queries  Machine learning: Small data, Complex models  CS Theory: (Randomized) Algorithms  Different cultures:  To a DB person, data mining is an extreme form of analytic processing – queries that examine large amounts of data  Result is the query answer  To a ML person, data-mining is the inference of models  Result is the parameters of the model  In this class we will do both! Machine Learning CS Theory Data Mining Database systems 11

12.  This class overlaps with machine learning, statistics, artificial intelligence, databases but more stress on  Scalability (big data)  Algorithms  Computing architectures  Automation for handling large data 12 Machine Learning Statistics Data Mining Database systems

13.  We will learn to mine different types of data:  Data is high dimensional  Data is a graph  Data is infinite/never-ending  Data is labeled  We will learn to use different models of computation:  MapReduce  Streams and online algorithms  Single machine in-memory 13

14.  We will learn to solve real-world problems:  Recommender systems  Market Basket Analysis  Spam detection  Duplicate document detection  We will learn various “tools”:  Linear algebra (SVD, Rec. Sys., Communities)  Optimization (stochastic gradient descent)  Dynamic programming (frequent itemsets)  Hashing (LSH, Bloom filters) 14

15. High dim. data Locality sensitive hashing Clustering Dimensional ity reduction Graph data PageRank, SimRank Community Detection Spam Detection Infinite data Filtering data streams Web advertising Queries on streams Machine learning SVM Decision Trees Perceptron, kNN Apps Recommen der systems Association Rules Duplicate document detection 15

16. 16 I ♥ data How do you want that data?

18.  Professor  Quinn Snell  3366 TMCB  801-422-5098  snell@cs.byu.edu  Office hours:  By Appointment 18

19.  Algorithms (CS312)  Dynamic programming, basic data structures  Basic probability  Moments, typical distributions, MLE, …  Programming (CS240)  Your choice, but C++/Java will be very useful  We provide some background, but the class will be fast paced 19

20.  Course website: http://bigdatalab.cs.byu.edu/cs476  Lecture slides  Homeworks, solutions  Readings  Readings:  Mining of Massive Datasets  Free online: http://www.mmds.orghttp://www.mmds.org  CRISP-DM 1.0 by Pete Chapman et. al.  ftp://ftp.software.ibm.com/software/analytics/spss/support/ Modeler/Documentation/14/UserManual/CRISP-DM.pdf 20

21.  For e-mailing assignments, always use:  bigdata.snell@gmail.com bigdata.snell@gmail.com  How to submit?  Homework write-up:  1-2 pages  Email PDF  Reading Quizzes  Noted on schedule with a Q 21

22.  Short reading quizzes: 5%  Midterm & Final exams: 40%  Homework & Labs: 55%  It’s going to be fun and hard work. 22

23.  I have big-data RA positions open!  I will post details 23