International Business Machines Corporation

Sap Se

Matthias Boehm

Berthold Reinwald

Shirish Tatikonda

Douglas Ronald Burdick

Shivakumar Vaithyanathan

Yuanyuan Tian

Prithviraj Sen

Alexandre Valentinovich Evfimievski

John D Keenleyside

Keith W Campbell

Arash Ashari

Robert Schulze

Wolfgang Lehner

Lars Dannecker

Mathias Peters

Stefan Burnicki

The embodiments described herein relate to recompiling an execution plan of a machine-learning program during runtime. An execution plan of a machine-learning program is compiled. In response to identifying a directed acyclic graph of high-level operations (HOP DAG) for recompilation during runtime, the execution plan is dynamically recompiled. The dynamic recompilation includes updating statistics and dynamically rewriting one or more operators of the identified HOP DAG, recomputing memory estimates of operators of the rewritten HOP DAG based on the updated statistics and rewritten operators, constructing a directed acyclic graph of low-level operations (LOP DAG) corresponding to the rewritten HOP DAG based in part on the recomputed memory estimates, and generating runtime instructions based on the LOP DAG.

Dynamic recompilation techniques for machine learning programs

A method for global data flow optimization for machine learning (ML) programs. The method includes receiving, by a storage device, an initial plan for an ML program. A processor builds a nested global data flow graph representation using the initial plan. Operator directed acyclic graphs (DAGs) are connected using crossblock operators according to inter-block data dependencies. The initial plan for the ML program is re-written resulting in an optimized plan for the ML program with respect to its global data flow properties. The re-writing includes re-writes of: configuration dataflow properties, operator selection and structural changes.

Global data flow optimization for machine learning programs

Parallel execution of machine learning programs is provided. Program code is received. The program code contains at least one parallel for statement having a plurality of iterations. A parallel execution plan is determined for the program code. According to the parallel execution plan, the plurality of iterations is partitioned into a plurality of tasks. Each task comprises at least one iteration. The iterations of each task are independent. Data required by the plurality of tasks is determined. An access pattern by the plurality of tasks of the data is determined. The data is partitioned based on the access pattern.

Hybrid parallelization strategies for machine learning programs on top of mapreduce

A method for optimization of machine learning (ML) workloads on a graphics processor unit (GPU). The method includes identifying a computation having a generic pattern commonly observed in ML processes. Hierarchical aggregation spanning a memory hierarchy of the GPU for processing is performed for the identified computation including maintaining partial output vector results in shared memory of the GPU. Hierarchical aggregation for vectors is performed including performing intra-block aggregation for multiple thread blocks of a partial output vector results on GPU global memory.

Pipelined approach to fused kernels for optimization of machine learning workloads on graphical processing units

One embodiment provides a method for runtime piggybacking of concurrent data-parallel jobs in task-parallel machine learning (ML) programs including intercepting, by a processor, executable jobs including executable map reduce (MR) jobs and looped jobs in a job stream. The processor queues the executable jobs, and applies runtime piggybacking of multiple jobs by processing workers of different types. Runtime piggybacking for a ParFOR (parallel for) ML program is optimized including configuring the runtime piggybacking based on processing worker type, degree of parallelism and minimum time thresholds.

Runtime piggybacking of concurrent jobs in task-parallel machine learning programs

A method for optimization of machine learning (ML) workloads on a graphics processor unit (GPU). The method includes identifying a computation having a generic pattern commonly observed in ML processes. An optimized fused GPU kernel is employed to exploit temporal locality for inherent data-flow dependencies in the identified computation. Hierarchical aggregation spanning a memory hierarchy of the GPU for processing for the identified computation is performed. GPU kernel launch parameters are estimated following an analytical model that maximizes thread occupancy and minimizes atomic writes to GPU global memory.

In a method for analyzing a large data set using a statistical computing environment language operation, a processor generates code from the statistical computing environment language operation that can be understood by a software system for processing machine learning algorithms in a MapReduce environment. A processor transfers the code to the software system for processing machine learning algorithms in a MapReduce environment. A processor invokes execution of the code with the software system for processing machine learning algorithms in a MapReduce environment.

R-language integration with a declarative machine learning language

Methods, systems, and computer-readable storage media for providing at least one parameter for use with a forecast model. Implementations include actions of receiving a first context vector, the first context vector including a plurality of context attributes that describe a first context, retrieving a first parameter vector from a repository based on the first context vector, the repository electronically storing a plurality of parameter vector, each parameter vector being associated with a respective context and including one or more parameters, parameterizing the forecast model based on parameters provided in the first parameter vector to provide a parameterized forecast model, optimizing the parameterized forecast model to provide an optimized forecast model, and forecasting one or more values using the optimized forecast model.

Context-aware parameter estimation for forecast models

Hybrid parallelization strategies for machine learning programs on top of MapReduce are provided. In one embodiment, a method of and computer program product for parallel execution of machine learning programs are provided. Program code is received. The program code contains at least one parallel for statement having a plurality of iterations. A parallel execution plan is determined for the program code. According to the parallel execution plan, the plurality of iterations is partitioned into a plurality of tasks. Each task comprises at least one iteration. The iterations of each task are independent.

Growing community of inventors

San Jose, CA, United States of America

Matthias Boehm