Optimization via too much Randomization Why parallelizing like
- Slides: 32
Optimization via (too much? ) Randomization Why parallelizing like crazy and being lazy can be good Peter Richtarik
Optimization as Mountain Climbing
Optimization with Big Data = Extreme* Mountain Climbing * in a billion dimensional space on a foggy day
Big Data BIG Volume BIG Velocity BIG Variety digital images & videos transaction records government records health records defence internet activity (social media, wikipedia, . . . ) • scientific measurements (physics, climate models, . . . ) • • •
God’s Algorithm = Teleportation
If You Are Not a God. . . x 0 x 2 x 3 x 1
Randomized Parallel Coordinate Descent holy grail settle for this start
Arup (Truss Topology Design) Western General Hospital (Creutzfeldt-Jakob Disease) Royal Observatory (Optimal Planet Growth) Ministry of Defence dstl lab (Algorithms for Data Simplicity)
Optimization as Lock Breaking
A Lock with 4 Dials A function representing the “quality” of a combination x = (x 1, x 2, x 3, x 4) F(x) = F(x 1, x 2, x 3, x 4) Setup: Combination maximizing F opens the lock Optimization Problem: Find combination maximizing F
Optimization Algorithm
A System of Billion Locks with Shared Dials 1) Nodes in the graph correspond to dials Lock x 1 x 4 x 3 x 2 xn 2) Nodes in the graph also correspond to locks: each lock (=node) owns dials connected to it in the graph by an edge # dials = n = # locks
How do we Measure the Quality of a Combination? • Each lock j has its own quality function Fj depending on the dials it owns • However, it does NOT open when Fj is maximized • The system of locks opens when F = F 1 + F 2 +. . . + Fn is maximized F : Rn R
An Algorithm with (too much? ) Randomization 1) Randomly select a lock 2) Randomly select a dial belonging to the lock 3) Adjust the value on the selected dial based only on the info corresponding to the selected lock
Synchronous Parallelization Processor 1 Processor 2 Processor 3 J 1 IDLE J 7 time IDLE E T S A W J 4 J 2 J 5 J 8 IDLE L FU J 3 IDLE J 6 J 9 IDLE
Crazy (Lock-Free) Parallelization Processor 1 Processor 2 Processor 3 J 1 J 4 J 3 W O N time J 2 J 5 J 7 J 8 E T S A J 6 J 9
Crazy Parallelization
Crazy Parallelization
Crazy Parallelization
Crazy Parallelization
Theoretical Result # Processors Average # of dials common between 2 locks # Locks Average # dials in a lock
Computational Insights
Theory vs Reality
Why parallelizing like crazy and being lazy can be good? Parallelization Randomization • • Effectivity Tractability Efficiency Scalability (big data) Parallelism Distribution Asynchronicity
Optimization Methods for Big Data • Randomized Coordinate Descent – P. R. and M. Takac: Parallel coordinate descent methods for big data optimization, Ar. Xiv: 1212. 0873 [can solve a problem with 1 billion variables in 2 hours using 24 processors] • Stochastic (Sub) Gradient Descent – P. R. and M. Takac: Randomized lock-free methods for minimizing partially separable convex functions [can be applied to optimize an unknown function] • Both of the above M. Takac, A. Bijral, P. R. and N. Srebro: Mini-batch primal and dual methods for support vector machines, Ar. Xiv: 1303. xxxx
Final 2 Slides
Probability HPC Matrix Theory Tools Machine Learning
- Kahoot quantifiers
- Incrementalizing graph algorithms
- Högkonjuktur inflation
- How much
- How much is too much plagiarism
- I feel like i care too much
- Andreas carlsson bye bye bye
- Too broad and too narrow examples
- Too broad too narrow
- Too broad and too narrow examples
- Too broad and too narrow examples
- Just about right scale
- Here you are too foreign for home
- Too anointed to be disappointed meaning
- Exact matrix completion via convex optimization
- The world is too much with us imagery
- The world is too much with us
- Analysis of the world is too much with us
- Megan and ron ate too much and felt sick.
- Security is always too much until the day it is not enough
- Too much vertical angulation results in images that are
- Deacon process reaction
- Good pilgrim you do wrong
- Never too much of a good thing
- Cause and effect essay conclusion
- Where does nitrogen come from
- Too much compassion
- Merits of problem solving method
- What is literary present
- It is unwise to pay too much
- Too much gravity
- How food travels through the digestive system
- If i profane with my unworthiest hand