Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Visualizing and Clustering Life Science Applications in Parallel 

9,151 views

Published on

HiCOMB 2015 14th IEEE International Workshop on
High Performance Computational Biology at IPDPS 2015
Hyderabad, India. This talk covers parallel data analytics for bioinformatics. Messages are
Always run MDS. Gives insight into data and performance of machine learning
Leads to a data browser as GIS gives for spatial data
3D better than 2D
~20D better than MSA?
Clustering Observations
Do you care about quality or are you just cutting up space into parts
Deterministic Clustering always makes more robust
Continuous clustering enables hierarchy
Trimmed Clustering cuts off tails
Distinct O(N) and O(N2) algorithms
Use Conjugate Gradient

Published in: Technology

Visualizing and Clustering Life Science Applications in Parallel 

  1. 1. Visualizing and Clustering Life Science Applications in Parallel HiCOMB 2015 14th IEEE International Workshop on High Performance Computational Biology at IPDPS 2015 Hyderabad, India May 25 2015 Geoffrey Fox gcf@indiana.edu http://www.infomall.org School of Informatics and Computing Digital Science Center Indiana University Bloomington 5/25/2015 1
  2. 2. Deterministic Annealing Algorithms 25/25/2015
  3. 3. Some Motivation • Big Data requires high performance – achieve with parallel computing • Big Data sometimes requires robust algorithms as more opportunity to make mistakes • Deterministic annealing (DA) is one of better approaches to robust optimization and broadly applicable – Started as “Elastic Net” by Durbin for Travelling Salesman Problem TSP – Tends to remove local optima – Addresses overfitting – Much Faster than simulated annealing • Physics systems find true lowest energy state if you anneal i.e. you equilibrate at each temperature as you cool • Uses mean field approximation, which is also used in “Variational Bayes” and “Variational inference”5/25/2015 3
  4. 4. (Deterministic) Annealing • Find minimum at high temperature when trivial • Small change avoiding local minima as lower temperature • Typically gets better answers than standard libraries- R and Mahout • And can be parallelized and put on GPU’s etc. 45/25/2015
  5. 5. General Features of DA • In many problems, decreasing temperature is classic multiscale – finer resolution (√T is “just” distance scale) • In clustering √T is distance in space of points (and centroids), for MDS scale in mapped Euclidean space • T = ∞, all points are in same place – the center of universe • For MDS all Euclidean points are at center and distances are zero. For clustering, there is one cluster • As Temperature lowered there are phase transitions in clustering cases where clusters split – Algorithm determines whether split needed as second derivative matrix singular • Note DA has similar features to hierarchical methods and you do not have to specify a number of clusters; you need to specify a final distance scale 55/25/2015
  6. 6. Math of Deterministic Annealing • H() is objective function to be minimized as a function of parameters  (as in Stress formula given earlier for MDS) • Gibbs Distribution at Temperature T P() = exp( - H()/T) /  d exp( - H()/T) • Or P() = exp( - H()/T + F/T ) • Use the Free Energy combining Objective Function and Entropy F = < H - T S(P) > =  d {P()H+ T P() lnP()} • Simulated annealing performs these integrals by Monte Carlo • Deterministic annealing corresponds to doing integrals analytically (by mean field approximation) and is much much faster • Need to introduce a modified Hamiltonian for some cases so that integrals are tractable. Introduce extra parameters to be varied so that modified Hamiltonian matches original • In each case temperature is lowered slowly – say by a factor 0.95 to 0.9999 at each iteration 5/25/2015 6
  7. 7. Some Uses of Deterministic Annealing DA • Clustering improved K-means – Vectors: Rose (Gurewitz and Fox 1990 – 486 citations encouraged me to revisit) – Clusters with fixed sizes and no tails (Proteomics team at Broad) – No Vectors: Hofmann and Buhmann (Just use pairwise distances) • Dimension Reduction for visualization and analysis – Vectors: GTM Generative Topographic Mapping – No vectors SMACOF: Multidimensional Scaling) MDS (Just use pairwise distances) • Can apply to HMM & general mixture models (less study) – Gaussian Mixture Models – Probabilistic Latent Semantic Analysis with Deterministic Annealing DA-PLSA as alternative to Latent Dirichlet Allocation for finding “hidden factors” • Have scalable parallel versions of much of above – mainly Java • Many clustering methods – not clear what is best although DA pretty good and improves K-means at increased computing cost which is not always useful • DA clearly improves MDS which is ~only reliable” dimension reduction? 5/25/2015 7
  8. 8. Clusters v. Regions • In Lymphocytes clusters are distinct; DA useful • In Pathology, clusters divide space into regions and sophisticated methods like deterministic annealing are probably unnecessary 8 Pathology 54D Lymphocytes 4D 5/25/2015
  9. 9. Some Problems we are working on • Analysis of Mass Spectrometry data to find peptides by clustering peaks (Broad Institute/Hyderabad) – ~0.5 million points in 2 dimensions (one collection) -- ~ 50,000 clusters summed over charges • Metagenomics – 0.5 million (increasing rapidly) points NOT in a vector space – hundreds of clusters per sample – Apply MDS to Phylogenetic trees • Pathology Images >50 Dimensions • Social image analysis is in a highish dimension vector space – 10-50 million images; 1000 features per image; million clusters • Finding communities from network graphs coming from Social media contacts etc. 95/25/2015
  10. 10. 5/25/2015 10 MDS and Clustering on ~60K EMR
  11. 11. 5/25/2015 11 Colored by Sample – not clustered. Distances from vectors in word space. This 128 4mers for ~200K sequences
  12. 12. Examples of current DA algorithms: LCMS 125/25/2015
  13. 13. Background on LC-MS • Remarks of collaborators – Broad Institute/Hyderabad • Abundance of peaks in “label-free” LC-MS enables large-scale comparison of peptides among groups of samples. • In fact when a group of samples in a cohort is analyzed together, not only is it possible to “align” robustly or cluster the corresponding peaks across samples, but it is also possible to search for patterns or fingerprints of disease states which may not be detectable in individual samples. • This property of the data lends itself naturally to big data analytics for biomarker discovery and is especially useful for population-level studies with large cohorts, as in the case of infectious diseases and epidemics. • With increasingly large-scale studies, the need for fast yet precise cohort-wide clustering of large numbers of peaks assumes technical importance. • In particular, a scalable parallel implementation of a cohort-wide peak clustering algorithm for LC-MS-based proteomic data can prove to be a critically important tool in clinical pipelines for responding to global epidemics of infectious diseases like tuberculosis, influenza, etc. 135/25/2015
  14. 14. Proteomics 2D DA Clustering T= 25000 with 60 Clusters (will be 30,000 at T=0.025) 5/25/2015 14
  15. 15. The brownish triangles are “sponge” (soaks up trash) peaks outside any cluster. The colored hexagons are peaks inside clusters with the white hexagons being determined cluster center 15 Fragment of 30,000 Clusters 241605 Points 5/25/2015
  16. 16. Continuous Clustering • This is a very useful subtlety introduced by Ken Rose but not widely known although it greatly improves algorithm • Take a cluster k to be split into 2 with centers Y(k)A and Y(k)B with initial values Y(k)A = Y(k)B at original center Y(k) • Then typically if you make this change and perturb the Y(k)A and Y(k)B, they will return to starting position as F at stable minimum (positive eigenvalue) • But instability (the negative eigenvalue) can develop and one finds • Implement by adding arbitrary number p(k) of centers for each cluster Zi = k=1 K p(k) exp(-i(k)/T) and M step gives p(k) = C(k)/N • Halve p(k) at splits; can’t split easily in standard case p(k) = 1 • Show weighting in sums like Zi now equipoint not equicluster as p(k) proportional to points C(k) in cluster 16 Free Energy F Y(k)A and Y(k)B Y(k)A + Y(k)B Free Energy F Free Energy F Y(k)A - Y(k)B 5/25/2015
  17. 17. Trimmed Clustering • Clustering with position-specific constraints on variance: Applying redescending M-estimators to label-free LC-MS data analysis (Rudolf Frühwirth, D R Mani and Saumyadipta Pyne) BMC Bioinformatics 2011, 12:358 • HTCC = k=0 K i=1 N Mi(k) f(i,k) – f(i,k) = (X(i) - Y(k))2/2(k)2 k > 0 – f(i,0) = c2 / 2 k = 0 • The 0’th cluster captures (at zero temperature) all points outside clusters (background) • Clusters are trimmed (X(i) - Y(k))2/2(k)2 < c2 / 2 • Relevant when well defined errors T ~ 0 T = 1 T = 5 Distance from cluster center 5/25/2015 17
  18. 18. 0 10000 20000 30000 40000 50000 60000 1.00E-031.00E-021.00E-011.00E+001.00E+011.00E+021.00E+031.00E+041.00E+051.00E+06 ClusterCount Temperature DAVS(2) DA2D Start Sponge DAVS(2) Add Close Cluster Check Sponge Reaches final value Cluster Count v. Temperature for 2 Runs • All start with one cluster at far left • T=1 special as measurement errors divided out • DA2D counts clusters with 1 member as clusters. DAVS(2) does not5/25/2015 18
  19. 19. Simple Parallelism as in k-means • Decompose points i over processors • Equations either pleasingly parallel “maps” over i • Or “All-Reductions” summing over i for each cluster • Parallel Algorithm: – Each process holds all clusters and calculates contributions to clusters from points in node – e.g. Y(k) = i=1 N <Mi(k)> Xi / C(k) • Runs well in MPI or MapReduce –See all the MapReduce k-means papers 195/25/2015
  20. 20. Better Parallelism • The previous model is correct at start but each point does not really contribute to each cluster as damped exponentially by exp( - (Xi- Y(k))2 /T ) • For Proteomics problem, on average only 6.45 clusters needed per point if require (Xi- Y(k))2 /T ≤ ~40 (as exp(-40) small) • So only need to keep nearby clusters for each point • As average number of Clusters ~ 20,000, this gives a factor of ~3000 improvement • Further communication is no longer all global; it has nearest neighbor components and calculated by parallelism over clusters which can be done in parallel if separated 205/25/2015
  21. 21. 21 Speedups for several runs on Tempest from 8-way through 384 way MPI parallelism with one thread per process. We look at different choices for MPI processes which are either inside nodes or on separate nodes 5/25/2015
  22. 22. 22 Parallelism within a Single Node of Madrid Cluster. A set of runs on 241605 peak data with a single node with 16 cores with either threads or MPI giving parallelism. Parallelism is either number of threads or number of MPI processes. Parallelism (#threads or #processes) 5/25/2015
  23. 23. METAGENOMICS -- SEQUENCE CLUSTERING Non-metric Spaces O(N2) Algorithms – Illustrate Phase Transitions 5/25/2015 23
  24. 24. 24 • Start at T= “” with 1 Cluster • Decrease T, Clusters emerge at instabilities 5/25/2015
  25. 25. 255/25/2015
  26. 26. 265/25/2015
  27. 27. 446K sequences ~100 clusters 5/25/2015 27
  28. 28. METAGENOMICS -- SEQUENCE CLUSTERING Non-metric Spaces O(N2) Algorithms – Compare Other Methods 5/25/2015 28
  29. 29. “Divergent” Data Sample 23 True Clusters 29 CDhit UClust Divergent Data Set UClust (Cuts 0.65 to 0.95) DAPWC 0.65 0.75 0.85 0.95 Total # of clusters 23 4 10 36 91 Total # of clusters uniquely identified 23 0 0 13 16 (i.e. one original cluster goes to 1 uclust cluster ) Total # of shared clusters with significant sharing 0 4 10 5 0 (one uclust cluster goes to > 1 real cluster) Total # of uclust clusters that are just part of a real cluster 0 4 10 17(11) 72(62) (numbers in brackets only have one member) Total # of real clusters that are 1 uclust cluster 0 14 9 5 0 but uclust cluster is spread over multiple real clusters Total # of real clusters that have 0 9 14 5 7 significant contribution from > 1 uclust cluster DA-PWC 5/25/2015
  30. 30. PROTEOMICS No clear clusters 5/25/2015 30
  31. 31. Protein Universe Browser for COG Sequences with a few illustrative biologically identified clusters 315/25/2015
  32. 32. Heatmap of biology distance (Needleman- Wunsch) vs 3D Euclidean Distances 32 If d a distance, so is f(d) for any monotonic f. Optimize choice of f 5/25/2015
  33. 33. O(N2) ALGORITHMS? 5/25/2015 33
  34. 34. Algorithm Challenges • See NRC Massive Data Analysis report • O(N) algorithms for O(N2) problems • Parallelizing Stochastic Gradient Descent • Streaming data algorithms – balance and interplay between batch methods (most time consuming) and interpolative streaming methods • Graph algorithms – need shared memory? • Machine Learning Community uses parameter servers; Parallel Computing (MPI) would not recommend this? – Is classic distributed model for “parameter service” better? • Apply best of parallel computing – communication and load balancing – to Giraph/Hadoop/Spark • Are data analytics sparse?; many cases are full matrices • BTW Need Java Grande – Some C++ but Java most popular in ABDS, with Python, Erlang, Go, Scala (compiles to JVM) …..5/25/2015 34
  35. 35. O(N2) interactions between green and purple clusters should be able to represent by centroids as in Barnes-Hut. Hard as no Gauss theorem; no multipole expansion and points really in 1000 dimension space as clustered before 3D projection O(N2) green-green and purple- purple interactions have value but green-purple are “wasted” “clean” sample of 446K 5/25/2015 35
  36. 36. 36 Use Barnes Hut OctTree, originally developed to make O(N2) astrophysics O(NlogN), to give similar speedups in machine learning 5/25/2015
  37. 37. 37 OctTree for 100K sample of Fungi We use OctTree for logarithmic interpolation (streaming data) 5/25/2015
  38. 38. Fungi Analysis 5/25/2015 38
  39. 39. Fungi Analysis • Multiple Species from multiple places • Several sources of sequences starting with 446K and eventually boiled down to ~10K curated sequences with 61 species • Original sample – clustering and MDS • Final sample – MDS and other clustering methods • Note MSA and SWG gives similar results • Some species are clearly split • Some species are diffuse; others compact making a fixed distance cut unreliable – Easy for humans! • MDS very clear on structure and clustering artifacts • Why not do “high-value” clustering as interactive iteration driven by MDS? 5/25/2015 39
  40. 40. Fungi -- 4 Classic Clustering Methods 5/25/2015 40
  41. 41. 5/25/2015 41
  42. 42. 5/25/2015 42
  43. 43. 5/25/2015 43 Same Species
  44. 44. 5/25/2015 44 Same Species Different Locations
  45. 45. Parallel Data Mining 5/25/2015 45
  46. 46. Parallel Data Analytics • Streaming algorithms have interesting differences but • “Batch” Data analytics is “just classic parallel computing” with usual features such as SPMD and BSP • Expect similar systematics to simulations where • Static Regular problems are straightforward but • Dynamic Irregular Problems are technically hard and high level approaches fail (see High Performance Fortran HPF) – Regular meshes worked well but – Adaptive dynamic meshes did not although “real people with MPI” could parallelize • However using libraries is successful at either – Lowest: communication level – Higher: “core analytics” level • Data analytics does not yet have “good regular parallel libraries” – Graph analytics has most attention 5/25/2015 46
  47. 47. Remarks on Parallelism I • Maximum Likelihood or 2 both lead to objective functions like • Minimize sum items=1 N (Positive nonlinear function of unknown parameters for item i) • Typically decompose items i and parallelize over both i and parameters to be determined • Solve iteratively with (clever) first or second order approximation to shift in objective function – Sometimes steepest descent direction; sometimes Newton – Have classic Expectation Maximization structure – Steepest descent shift is sum over shift calculated from each point • Classic method – take all (millions) of items in data set and move full distance – Stochastic Gradient Descent SGD – take randomly a few hundred of items in data set and calculate shifts over these and move a tiny distance – SGD cannot parallelize over items 475/25/2015
  48. 48. Remarks on Parallelism II • Need to cover non vector semimetric and vector spaces for clustering and dimension reduction (N points in space) • Semimetric spaces just have pairwise distances defined between points in space (i, j) • MDS Minimizes Stress and illustrates this (X) = i<j=1 N weight(i,j) ((i, j) - d(Xi , Xj))2 • Vector spaces have Euclidean distance and scalar products – Algorithms can be O(N) and these are best for clustering but for MDS O(N) methods may not be best as obvious objective function O(N2) – Important new algorithms needed to define O(N) versions of current O(N2) – “must” work intuitively and shown in principle • Note matrix solvers often use conjugate gradient – converges in 5- 100 iterations – a big gain for matrix with a million rows. This removes factor of N in time complexity • Ratio of #clusters to #points important; new clustering ideas if ratio >~ 0.1 485/25/2015
  49. 49. Problem Structure • Note learning networks have huge number of parameters (11 billion in Stanford work) so that inconceivable to look at second derivative • Clustering and MDS have lots of parameters but can be practical to look at second derivative and use Newton’s method to minimize • Parameters are determined in distributed fashion but are typically needed globally – MPI use broadcast and “AllCollectives” implying Map-Collective is a useful programming model – AI community: use parameter server and access as needed. Non-optimal? 495/25/2015 (1) Map Only (4) Point to Point or Map-Communication (3) Iterative Map Reduce or Map-Collective (2) Classic MapReduce Input map reduce Input map reduce Iterations Input Output map Local Graph (5) Map-Streaming maps brokers Events (6) Shared memory Map Communicates Map & Communicate Shared Memory
  50. 50. MDS in more detail 5/25/2015 50
  51. 51. WDA-SMACOF “Best” MDS • Semimetric spaces just have pairwise distances defined between points in space (i, j) • MDS Minimizes Stress with pairwise distances (i, j) (X) = i<j=1 N weight(i,j) ((i, j) - d(Xi , Xj))2 • SMACOF clever Expectation Maximization method choses good steepest descent • Improved by Deterministic Annealing reducing distance scale; DA does not impact compute time much and gives DA-SMACOF • Classic SMACOF is O(N2) for uniform weight and O(N3) for non trivial weights but get nonuniform weight from – The preferred Sammon method weight(i,j) = 1/(i, j) or – Missing distances put in as weight(i,j) = 0 • Use conjugate gradient – converges in 5-100 iterations – a big gain for matrix with a million rows. This removes factor of N in time complexity and gives WDA-SMACOF 515/25/2015
  52. 52. Timing of WDA SMACOF • 20k to 100k AM Fungal sequences on 600 cores 5/25/2015 52 1 10 100 1000 10000 100000 20k 40k 60k 80k 100k Seconds Data Size Time Cost Comparison between WDA- SMACOF with Equal Weights and Sammon's Mapping Equal Weights Sammon's Mapping
  53. 53. WDA-SMACOF Timing • Input Data: 100k to 400k AM Fungal sequences • Environment: 32 nodes (1024 cores) to 128 nodes (4096 cores) on BigRed2. • Using Harp plug in for Hadoop (MPI Performance) 0 500 1000 1500 2000 2500 3000 3500 4000 100k 200k 300k 400k Seconds Data Size Time Cost of WDA-SMACOF over Increasing Data Size 512 1024 2048 4096 0 0.2 0.4 0.6 0.8 1 1.2 512 1024 2048 4096 ParallelEfficiency Number of Processors Parallel Efficiency of WDA-SMACOF over Increasing Number of Processors WDA-SMACOF (Harp) 5/25/2015 53
  54. 54. Spherical Phylogram • Take a set of sequences mapped to nD with MDS (WDA- SMACOF) (n=3 or ~20) – N=20 captures ~all features of dataset? • Consider a phylogenetic tree and use neighbor joining formulae to calculate distances of nodes to sequences (or later other nodes) starting at bottom of tree • Do a new MDS fixing mapping of sequences noting that sequences + nodes have defined distances • Use RAxML or Neighbor Joining (N=20?) to find tree • Random note: do not need Multiple Sequence Alignment; pairwise tools are easier to use and give reliably good results 5/25/2015 54
  55. 55. RAxML result visualized in FigTree. Spherical Phylogram visualized in PlotViz for MSA or SWG distances Spherical Phylograms MSA SWG 5/25/2015 55
  56. 56. Quality of 3D Phylogenetic Tree • EM-SMACOF is basic SMACOF • LMA was previous best method using Levenberg-Marquardt nonlinear 2 solver • WDA-SMACOF finds best result • 3 different distance measures Sum of branch lengths of the Spherical Phylogram generated in 3D space on two datasets 0 5 10 15 20 25 30 MSA SWG NW SumofBranches Sum of Branches on 599nts Data WDA-SMACOF LMA EM-SMACOF 0 5 10 15 20 25 MSA SWG NW SumofBranches Sum of Branches on 999nts Data WDA-SMACOF LMA EM-SMACOF 5/25/2015 56
  57. 57. Summary • Always run MDS. Gives insight into data and performance of machine learning – Leads to a data browser as GIS gives for spatial data – 3D better than 2D – ~20D better than MSA? • Clustering Observations – Do you care about quality or are you just cutting up space into parts – Deterministic Clustering always makes more robust – Continuous clustering enables hierarchy – Trimmed Clustering cuts off tails – Distinct O(N) and O(N2) algorithms • Use Conjugate Gradient 575/25/2015
  58. 58. Java Grande 5/25/2015 58
  59. 59. Java Grande • We once tried to encourage use of Java in HPC with Java Grande Forum but Fortran, C and C++ remain central HPC languages. – Not helped by .com and Sun collapse in 2000-2005 • The pure Java CartaBlanca, a 2005 R&D100 award-winning project, was an early successful example of HPC use of Java in a simulation tool for non-linear physics on unstructured grids. • Of course Java is a major language in ABDS and as data analysis and simulation are naturally linked, should consider broader use of Java • Using Habanero Java (from Rice University) for Threads and mpiJava or FastMPJ for MPI, gathering collection of high performance parallel Java analytics – Converted from C# and sequential Java faster than sequential C# • So will have either Hadoop+Harp or classic Threads/MPI versions in Java Grande version of Mahout5/25/2015 59
  60. 60. Performance of MPI Kernel Operations 1 100 10000 0B 2B 8B 32B 128B 512B 2KB 8KB 32KB 128KB 512KB Averagetime(us) Message size (bytes) MPI.NET C# in Tempest FastMPJ Java in FG OMPI-nightly Java FG OMPI-trunk Java FG OMPI-trunk C FG Performance of MPI send and receive operations 5 5000 4B 16B 64B 256B 1KB 4KB 16KB 64KB 256KB 1MB 4MB Averagetime(us) Message size (bytes) MPI.NET C# in Tempest FastMPJ Java in FG OMPI-nightly Java FG OMPI-trunk Java FG OMPI-trunk C FG Performance of MPI allreduce operation 1 100 10000 1000000 4B 16B 64B 256B 1KB 4KB 16KB 64KB 256KB 1MB 4MB AverageTime(us) Message Size (bytes) OMPI-trunk C Madrid OMPI-trunk Java Madrid OMPI-trunk C FG OMPI-trunk Java FG 1 10 100 1000 10000 0B 2B 8B 32B 128B 512B 2KB 8KB 32KB 128KB 512KB AverageTime(us) Message Size (bytes) OMPI-trunk C Madrid OMPI-trunk Java Madrid OMPI-trunk C FG OMPI-trunk Java FG Performance of MPI send and receive on Infiniband and Ethernet Performance of MPI allreduce on Infiniband and Ethernet Pure Java as in FastMPJ slower than Java interfacing to C version of MPI 5/25/2015 60
  61. 61. Java Grande and C# on 40K point DAPWC Clustering Very sensitive to threads v MPI 64 Way parallel 128 Way parallel 256 Way parallel TXP Nodes Total C# Java C# Hardware 0.7 performance Java Hardware 5/25/2015 61
  62. 62. Java and C# on 12.6K point DAPWC Clustering Java C##Threads x #Processes per node # Nodes Total Parallelism Time hours 1x1 2x21x2 1x42x1 1x84x1 2x4 4x2 8x1 #Threads x #Processes per node C# Hardware 0.7 performance Java Hardware 5/25/2015 62
  63. 63. Data Analytics in SPIDAL 5/25/2015 63
  64. 64. Analytics and the DIKW Pipeline • Data goes through a pipeline Raw data  Data  Information  Knowledge  Wisdom  Decisions • Each link enabled by a filter which is “business logic” or “analytics” • We are interested in filters that involve “sophisticated analytics” which require non trivial parallel algorithms – Improve state of art in both algorithm quality and (parallel) performance • Design and Build SPIDAL (Scalable Parallel Interoperable Data Analytics Library) More Analytics KnowledgeInformation Analytics InformationData 5/25/2015 64
  65. 65. Strategy to Build SPIDAL • Analyze Big Data applications to identify analytics needed and generate benchmark applications • Analyze existing analytics libraries (in practice limit to some application domains) – catalog library members available and performance – Mahout low performance, R largely sequential and missing key algorithms, MLlib just starting • Identify big data computer architectures • Identify software model to allow interoperability and performance • Design or identify new or existing algorithm including parallel implementation • Collaborate application scientists, computer systems and statistics/algorithms communities5/25/2015 65
  66. 66. Machine Learning in Network Science, Imaging in Computer Vision, Pathology, Polar Science, Biomolecular Simulations 66 Algorithm Applications Features Status Parallelism Graph Analytics Community detection Social networks, webgraph Graph . P-DM GML-GrC Subgraph/motif finding Webgraph, biological/social networks P-DM GML-GrB Finding diameter Social networks, webgraph P-DM GML-GrB Clustering coefficient Social networks P-DM GML-GrC Page rank Webgraph P-DM GML-GrC Maximal cliques Social networks, webgraph P-DM GML-GrB Connected component Social networks, webgraph P-DM GML-GrB Betweenness centrality Social networks Graph, Non-metric, static P-Shm GML-GRA Shortest path Social networks, webgraph P-Shm Spatial Queries and Analytics Spatial relationship based queries GIS/social networks/pathology informatics Geometric P-DM PP Distance based queries P-DM PP Spatial clustering Seq GML Spatial modeling Seq PP GML Global (parallel) ML GrA Static GrB Runtime partitioning 5/25/2015
  67. 67. Some specialized data analytics in SPIDAL • aa 67 Algorithm Applications Features Status Parallelism Core Image Processing Image preprocessing Computer vision/pathology informatics Metric Space Point Sets, Neighborhood sets & Image features P-DM PP Object detection & segmentation P-DM PP Image/object feature computation P-DM PP 3D image registration Seq PP Object matching Geometric Todo PP 3D feature extraction Todo PP Deep Learning Learning Network, Stochastic Gradient Descent Image Understanding, Language Translation, Voice Recognition, Car driving Connections in artificial neural net P-DM GML PP Pleasingly Parallel (Local ML) Seq Sequential Available GRA Good distributed algorithm needed Todo No prototype Available P-DM Distributed memory Available P-Shm Shared memory Available5/25/2015
  68. 68. Some Core Machine Learning Building Blocks 68 Algorithm Applications Features Status //ism DA Vector Clustering Accurate Clusters Vectors P-DM GML DA Non metric Clustering Accurate Clusters, Biology, Web Non metric, O(N2) P-DM GML Kmeans; Basic, Fuzzy and Elkan Fast Clustering Vectors P-DM GML Levenberg-Marquardt Optimization Non-linear Gauss-Newton, use in MDS Least Squares P-DM GML SMACOF Dimension Reduction DA- MDS with general weights Least Squares, O(N2) P-DM GML Vector Dimension Reduction DA-GTM and Others Vectors P-DM GML TFIDF Search Find nearest neighbors in document corpus Bag of “words” (image features) P-DM PP All-pairs similarity search Find pairs of documents with TFIDF distance below a threshold Todo GML Support Vector Machine SVM Learn and Classify Vectors Seq GML Random Forest Learn and Classify Vectors P-DM PP Gibbs sampling (MCMC) Solve global inference problems Graph Todo GML Latent Dirichlet Allocation LDA with Gibbs sampling or Var. Bayes Topic models (Latent factors) Bag of “words” P-DM GML Singular Value Decomposition SVD Dimension Reduction and PCA Vectors Seq GML Hidden Markov Models (HMM) Global inference on sequence models Vectors Seq PP & GML 5/25/2015
  69. 69. Some Futures • Always run MDS. Gives insight into data – Leads to a data browser as GIS gives for spatial data • Claim is algorithm change gave as much performance increase as hardware change in simulations. Will this happen in analytics? – Today is like parallel computing 30 years ago with regular meshs. We will learn how to adapt methods automatically to give “multigrid” and “fast multipole” like algorithms • Need to start developing the libraries that support Big Data – Understand architectures issues – Have coupled batch and streaming versions – Develop much better algorithms • Please join SPIDAL (Scalable Parallel Interoperable Data Analytics Library) community 695/25/2015

×