[jira] [Created] (IGNITE-5000) Rename Ignite Math module to Ignite ML module
Yury Babak created IGNITE-5000: -- Summary: Rename Ignite Math module to Ignite ML module Key: IGNITE-5000 URL: https://issues.apache.org/jira/browse/IGNITE-5000 Project: Ignite Issue Type: Task Affects Versions: 2.0 Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.0 We want to rename math module to ml(machine learning) because it's a main planned functionality for this module. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5029) Ignite ML next steps
Yury Babak created IGNITE-5029: -- Summary: Ignite ML next steps Key: IGNITE-5029 URL: https://issues.apache.org/jira/browse/IGNITE-5029 Project: Ignite Issue Type: Task Reporter: Yury Babak Assignee: Yury Babak We want to find out what should be done next 3 months. Currently we thinking about the following areas: 1) [~oignatenko] - Regression and clustering algorithms. 2) [~amalykh] - Deep Learning/Neural Networks stuff. 3) [~chief] - DSL/scripting support. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5880) BLAS integration phase 2
Yury Babak created IGNITE-5880: -- Summary: BLAS integration phase 2 Key: IGNITE-5880 URL: https://issues.apache.org/jira/browse/IGNITE-5880 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Yury Babak Fix For: 2.2 The second phase of BLAS integration. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5925) Get row/col for matrices
Yury Babak created IGNITE-5925: -- Summary: Get row/col for matrices Key: IGNITE-5925 URL: https://issues.apache.org/jira/browse/IGNITE-5925 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.2 It should be useful to have this api for any matrix especially in BLAS and decompositions. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6059) Use any distributed matrix in K-Means
Yury Babak created IGNITE-6059: -- Summary: Use any distributed matrix in K-Means Key: IGNITE-6059 URL: https://issues.apache.org/jira/browse/IGNITE-6059 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.2 Currently k-means work only with row/col matrix. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5777) BLAS integration
Yury Babak created IGNITE-5777: -- Summary: BLAS integration Key: IGNITE-5777 URL: https://issues.apache.org/jira/browse/IGNITE-5777 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak Replace all naive computations by BLAS from 5278 -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5724) Remove all autoboxing staff from the component.
Yury Babak created IGNITE-5724: -- Summary: Remove all autoboxing staff from the component. Key: IGNITE-5724 URL: https://issues.apache.org/jira/browse/IGNITE-5724 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Find and remove all boxing/unboxing code from vectors and matrices. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5725) Adopt TracerTest for TC builds.
Yury Babak created IGNITE-5725: -- Summary: Adopt TracerTest for TC builds. Key: IGNITE-5725 URL: https://issues.apache.org/jira/browse/IGNITE-5725 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Currently TracerTest runs browser but we don`t have even X11 on TC. So we need to decide - run add possibility to run browser on TC, change tests and don`t call "Desktop.getDesktop().browse" during TC build or just ignore this test for TC build. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5723) Improve code quality for existing code.
Yury Babak created IGNITE-5723: -- Summary: Improve code quality for existing code. Key: IGNITE-5723 URL: https://issues.apache.org/jira/browse/IGNITE-5723 Project: Ignite Issue Type: Improvement Components: ml Environment: (?) check code style for all sources. (?) check code coverage. Reporter: Yury Babak -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5791) Block matrix Introduction
Yury Babak created IGNITE-5791: -- Summary: Block matrix Introduction Key: IGNITE-5791 URL: https://issues.apache.org/jira/browse/IGNITE-5791 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.2 Introduction of sparce distributed block matrix with more efficient multiplication rather than row/col distributed matrix. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5801) Externalization for offheap vectors/matrices
Yury Babak created IGNITE-5801: -- Summary: Externalization for offheap vectors/matrices Key: IGNITE-5801 URL: https://issues.apache.org/jira/browse/IGNITE-5801 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Add externalization support for off-heap structures. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5825) K-Means refactoring
Yury Babak created IGNITE-5825: -- Summary: K-Means refactoring Key: IGNITE-5825 URL: https://issues.apache.org/jira/browse/IGNITE-5825 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Improve performance of points copying. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5824) Adjust precision in math unit tests.
Yury Babak created IGNITE-5824: -- Summary: Adjust precision in math unit tests. Key: IGNITE-5824 URL: https://issues.apache.org/jira/browse/IGNITE-5824 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Find which precision is sufficient for math related tests. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5827) Benchmarks refactoring
Yury Babak created IGNITE-5827: -- Summary: Benchmarks refactoring Key: IGNITE-5827 URL: https://issues.apache.org/jira/browse/IGNITE-5827 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak See MathBenchmark.java and VectorBenchmarkTest.java -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5826) OLS refactoring
Yury Babak created IGNITE-5826: -- Summary: OLS refactoring Key: IGNITE-5826 URL: https://issues.apache.org/jira/browse/IGNITE-5826 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak (?) Copying. (?) Incremental formula. (?) Tests. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5828) Decompositions refactoring
Yury Babak created IGNITE-5828: -- Summary: Decompositions refactoring Key: IGNITE-5828 URL: https://issues.apache.org/jira/browse/IGNITE-5828 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak (?) Externalization for decompositions. (?) QRDecomposition performance. (?) EigenDecompositionTest - corner case failure. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5856) BLAS integration phase 1
Yury Babak created IGNITE-5856: -- Summary: BLAS integration phase 1 Key: IGNITE-5856 URL: https://issues.apache.org/jira/browse/IGNITE-5856 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Yury Babak (i) BLAS multiplication for dense and sparse local matrices. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5845) Benchmarks for ML algorithms.
Yury Babak created IGNITE-5845: -- Summary: Benchmarks for ML algorithms. Key: IGNITE-5845 URL: https://issues.apache.org/jira/browse/IGNITE-5845 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak We want to create some benchmarks for ML algorithms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5844) Distributed versions of matrix decompositions
Yury Babak created IGNITE-5844: -- Summary: Distributed versions of matrix decompositions Key: IGNITE-5844 URL: https://issues.apache.org/jira/browse/IGNITE-5844 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak We want to add support for distributed matrices. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5846) Add support of distributed matrices for OLS regression.
Yury Babak created IGNITE-5846: -- Summary: Add support of distributed matrices for OLS regression. Key: IGNITE-5846 URL: https://issues.apache.org/jira/browse/IGNITE-5846 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Currently OSL regression works only with local matrices. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5847) API for Models
Yury Babak created IGNITE-5847: -- Summary: API for Models Key: IGNITE-5847 URL: https://issues.apache.org/jira/browse/IGNITE-5847 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak We want to design and implement API for Models. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5799) Caching for some intermediate calcs
Yury Babak created IGNITE-5799: -- Summary: Caching for some intermediate calcs Key: IGNITE-5799 URL: https://issues.apache.org/jira/browse/IGNITE-5799 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Check possibility and necessity of caching some intermediate calcs like decomposition for matrix determinant calculation -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5802) Fix all wrong TODO comments in component
Yury Babak created IGNITE-5802: -- Summary: Fix all wrong TODO comments in component Key: IGNITE-5802 URL: https://issues.apache.org/jira/browse/IGNITE-5802 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.2 (i) https://cwiki.apache.org/confluence/display/IGNITE/Coding+Guidelines#CodingGuidelines-TODOs Not all TODOs in componet following those rules, that must be fixed. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5645) Locking mechanism for distributed matrices.
Yury Babak created IGNITE-5645: -- Summary: Locking mechanism for distributed matrices. Key: IGNITE-5645 URL: https://issues.apache.org/jira/browse/IGNITE-5645 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.2 We must to have mechanism for protect distributed matrix from changes during calculations. Current locking mechanism is bad choice for locking a huge cache keyset, so we need a new one. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5646) Use affinity keys for distributed matrice blocks
Yury Babak created IGNITE-5646: -- Summary: Use affinity keys for distributed matrice blocks Key: IGNITE-5646 URL: https://issues.apache.org/jira/browse/IGNITE-5646 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.2 We want to implement affinity collocation for distributed matrices. We must guarantee that the new block for computation result will be stored in the same node like the initial blocks -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-5065) DSL/scription support
Yury Babak created IGNITE-5065: -- Summary: DSL/scription support Key: IGNITE-5065 URL: https://issues.apache.org/jira/browse/IGNITE-5065 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak The goal is introduce JS(using Nashorn) support as scripting language. Also we should make investigation about using Scala as DSL. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5112) Create OLS example
Yury Babak created IGNITE-5112: -- Summary: Create OLS example Key: IGNITE-5112 URL: https://issues.apache.org/jira/browse/IGNITE-5112 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak Fix For: 2.1 We want to have example of using OLS algorithm. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5113) K-means clustering
Yury Babak created IGNITE-5113: -- Summary: K-means clustering Key: IGNITE-5113 URL: https://issues.apache.org/jira/browse/IGNITE-5113 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Artem Malykh Fix For: 2.1 -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5114) Parallalel plus/multiply matrix algorithms.
Yury Babak created IGNITE-5114: -- Summary: Parallalel plus/multiply matrix algorithms. Key: IGNITE-5114 URL: https://issues.apache.org/jira/browse/IGNITE-5114 Project: Ignite Issue Type: Improvement Reporter: Yury Babak We need design and implement parallel(distributed?) versions of matrix plus/multiply algorithms because current implementation is implemented by definition. Those implementations is really slow. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5109) Refactoring for SparseDistributedMatrix
Yury Babak created IGNITE-5109: -- Summary: Refactoring for SparseDistributedMatrix Key: IGNITE-5109 URL: https://issues.apache.org/jira/browse/IGNITE-5109 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.1 According to [~agoncharuk] we should change logic of this matrix. Creating new cache each time when we create new matrix is really bad idea. So we must use one dedicated cache for ML matrices. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5216) Add Stream API support to Ignite ML matrices.
Yury Babak created IGNITE-5216: -- Summary: Add Stream API support to Ignite ML matrices. Key: IGNITE-5216 URL: https://issues.apache.org/jira/browse/IGNITE-5216 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak We want to add Stream API support to Ignite matrices and possibly to vectors. We already have implementation of Spliterator for AbstractVector and AbstractMatrix so it's looks like next step. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5217) Gradient descent
Yury Babak created IGNITE-5217: -- Summary: Gradient descent Key: IGNITE-5217 URL: https://issues.apache.org/jira/browse/IGNITE-5217 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak We should implement Gradient Descent for Ignite ML. Current we have only normal equation for linear regression. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5220) Partial derivatives calculation.
Yury Babak created IGNITE-5220: -- Summary: Partial derivatives calculation. Key: IGNITE-5220 URL: https://issues.apache.org/jira/browse/IGNITE-5220 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak We need mechanism of computation of partial derivatives which we need for gradient descent. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5218) Decision trees
Yury Babak created IGNITE-5218: -- Summary: Decision trees Key: IGNITE-5218 URL: https://issues.apache.org/jira/browse/IGNITE-5218 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak We want to implement Decision trees for Ignite ML because it's really common one for ML. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5246) Fuzzy c-means (FCM)
Yury Babak created IGNITE-5246: -- Summary: Fuzzy c-means (FCM) Key: IGNITE-5246 URL: https://issues.apache.org/jira/browse/IGNITE-5246 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak We want to have implementation of fuzzy clusterization. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5280) SparseDistributedMatrix refactorig
Yury Babak created IGNITE-5280: -- Summary: SparseDistributedMatrix refactorig Key: IGNITE-5280 URL: https://issues.apache.org/jira/browse/IGNITE-5280 Project: Ignite Issue Type: Sub-task Reporter: Yury Babak Assignee: Yury Babak We must refactor SparseDistributedMatrix for decrease communication during computations. -- This message was sent by Atlassian JIRA (v6.3.15#6346)
[jira] [Created] (IGNITE-5535) BLAS support for offheap vector/matrix
Yury Babak created IGNITE-5535: -- Summary: BLAS support for offheap vector/matrix Key: IGNITE-5535 URL: https://issues.apache.org/jira/browse/IGNITE-5535 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak We want to add BLAS support for offheap stuctures. Current we implement only onheap version. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6373) Create example for local and distributed k-means algorithm
Yury Babak created IGNITE-6373: -- Summary: Create example for local and distributed k-means algorithm Key: IGNITE-6373 URL: https://issues.apache.org/jira/browse/IGNITE-6373 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.3 Currently we no examples for both versions of k-means. So we need at least two example for local k-means and for distributed k-means. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6386) Introduction of distributed neural networks.
Yury Babak created IGNITE-6386: -- Summary: Introduction of distributed neural networks. Key: IGNITE-6386 URL: https://issues.apache.org/jira/browse/IGNITE-6386 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.3 We want to have deep learning algorithms and for this we need to implement neural network over Apache Ignite. Currently we think about using [dl4j|https://deeplearning4j.org] as backend but in this case we cannot train one model over multiple nodes efficiently. Also we will think about integration/connector with dl4j. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6585) SVM for Apache Ignite ML module
Yury Babak created IGNITE-6585: -- Summary: SVM for Apache Ignite ML module Key: IGNITE-6585 URL: https://issues.apache.org/jira/browse/IGNITE-6585 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak SVM - support vector machine, is pretty common algorithm and I think that we need it in our module. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6642) Integration with PMML
Yury Babak created IGNITE-6642: -- Summary: Integration with PMML Key: IGNITE-6642 URL: https://issues.apache.org/jira/browse/IGNITE-6642 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak PMML - Predictive Model Markup Language is XML based language which used in SPARK MLlib and others platforms. Here some additional info about PMML: (i) http://dmg.org/pmml/v4-3/GeneralStructure.html (i) https://github.com/jpmml/jpmml-model -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6640) Introduction of models import/export
Yury Babak created IGNITE-6640: -- Summary: Introduction of models import/export Key: IGNITE-6640 URL: https://issues.apache.org/jira/browse/IGNITE-6640 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak Assignee: Yury Babak We need to add basic import/export functionality for ml models. We will start from simple binary save to file and load from file. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6222) Adapt Linear regression for distributed enviroment
Yury Babak created IGNITE-6222: -- Summary: Adapt Linear regression for distributed enviroment Key: IGNITE-6222 URL: https://issues.apache.org/jira/browse/IGNITE-6222 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Fix For: 2.3 Currently our implementation of linear regression works only locally which is useless. So we need fix this implementation. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6497) Broken tests in ignite-2.1.5
Yury Babak created IGNITE-6497: -- Summary: Broken tests in ignite-2.1.5 Key: IGNITE-6497 URL: https://issues.apache.org/jira/browse/IGNITE-6497 Project: Ignite Issue Type: Bug Components: ml Affects Versions: 2.1, 2.2 Reporter: Yury Babak Assignee: Artem Malykh https://ci.ignite.apache.org/viewLog.html?buildId=852249=buildResultsDiv=Ignite20Tests_IgniteMl -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6123) First performance measurement of implemented ml algorithms
Yury Babak created IGNITE-6123: -- Summary: First performance measurement of implemented ml algorithms Key: IGNITE-6123 URL: https://issues.apache.org/jira/browse/IGNITE-6123 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.2 We want to start tracking our performance to avoid performance degradation. Also we need some performance comparison with other ml libs. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6880) KNN(k nearest neighbor) algorithm
Yury Babak created IGNITE-6880: -- Summary: KNN(k nearest neighbor) algorithm Key: IGNITE-6880 URL: https://issues.apache.org/jira/browse/IGNITE-6880 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak We want to add KNN to Apache Ignite ML module. Our implementation should support two modes: * KNN-classifier(in this mode the output is a class membership. An object is classified by a majority vote of its neighbors, with the object being assigned to the class most common among its k nearest neighbors (k is a positive integer, typically small). If k = 1, then the object is simply assigned to the class of that single nearest neighbor) * KNN-regression(the output is the property value for the object. This value is the average of the values of its k nearest neighbors.) -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6878) Naive Bayes classifier for ML module
Yury Babak created IGNITE-6878: -- Summary: Naive Bayes classifier for ML module Key: IGNITE-6878 URL: https://issues.apache.org/jira/browse/IGNITE-6878 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak Naive Bayes classifiers are a family of simple probabilistic classifiers based on applying Bayes' theorem with strong (naive) independence assumptions between the features. So we want to add this algorithm to Apache Ignite ML module. Ideally, implementation should support both multinomial naive Bayes and Bernoulli naive Bayes. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6884) Implement of tensorFold() and tensorProduct() for vectors and matrices
Yury Babak created IGNITE-6884: -- Summary: Implement of tensorFold() and tensorProduct() for vectors and matrices Key: IGNITE-6884 URL: https://issues.apache.org/jira/browse/IGNITE-6884 Project: Ignite Issue Type: Improvement Security Level: Public (Viewable by anyone) Reporter: Yury Babak We want to implement tensor fold and map for matrices and vectors. Also we must take into consideration the different types of matrices vectors, including distribution versions. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6882) Introduction of computation graph
Yury Babak created IGNITE-6882: -- Summary: Introduction of computation graph Key: IGNITE-6882 URL: https://issues.apache.org/jira/browse/IGNITE-6882 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 We want to implement a computation graph for NNs because this should helps us achieve for neural networks not only data parallelism but model parallelism too. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7205) DataSet API
Yury Babak created IGNITE-7205: -- Summary: DataSet API Key: IGNITE-7205 URL: https://issues.apache.org/jira/browse/IGNITE-7205 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak We want to create and implement API for Dataset. This should include Dataset, Labeled Dataset, dataset preprocessors(normalizer, filter, mapper, etc...) -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7213) Empty class descriptions for KNNModelFormat
Yury Babak created IGNITE-7213: -- Summary: Empty class descriptions for KNNModelFormat Key: IGNITE-7213 URL: https://issues.apache.org/jira/browse/IGNITE-7213 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Priority: Critical Fix For: 2.4 Javadoc generation failed if we have classes with empty class-javadoc -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7242) Broken javadoc for KNN
Yury Babak created IGNITE-7242: -- Summary: Broken javadoc for KNN Key: IGNITE-7242 URL: https://issues.apache.org/jira/browse/IGNITE-7242 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6899) Adding GA Grid to Apache Ignite ML module.
Yury Babak created IGNITE-6899: -- Summary: Adding GA Grid to Apache Ignite ML module. Key: IGNITE-6899 URL: https://issues.apache.org/jira/browse/IGNITE-6899 Project: Ignite Issue Type: New Feature Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak Fix For: 2.4 We want to add GA Grid to our ML Module. This is the first iteration of this integration. On this step we will simple add GA Grid to the separate package in ML module. (i) This is a good package for GA Grid: org.apache.ignite.ml.genetic (i) For GA Grid we need unit tests as well as examples -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6949) Cleanup OLS code
Yury Babak created IGNITE-6949: -- Summary: Cleanup OLS code Key: IGNITE-6949 URL: https://issues.apache.org/jira/browse/IGNITE-6949 Project: Ignite Issue Type: Bug Security Level: Public (Viewable by anyone) Reporter: Yury Babak Assignee: Aleksey Zinoviev Fix For: 2.4 We want fix wrong styles like wildcards in imports, unnecessary empty lines, missed empty lines and if-else blocks format in OLS related files. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-6849) Failed k-means tests with multiple nodes
Yury Babak created IGNITE-6849: -- Summary: Failed k-means tests with multiple nodes Key: IGNITE-6849 URL: https://issues.apache.org/jira/browse/IGNITE-6849 Project: Ignite Issue Type: Bug Security Level: Public (Viewable by anyone) Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 * MapWrapperVector: missed no-op constructor. * Cardinality violation exceptions. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7149) Gradient boosting for decision tree
Yury Babak created IGNITE-7149: -- Summary: Gradient boosting for decision tree Key: IGNITE-7149 URL: https://issues.apache.org/jira/browse/IGNITE-7149 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak We want to implement gradient boosting for decision trees. It should be new implementation of Trainer interface and we should keep possibility to choose which trainer we want to use for our tree. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7150) Gradient boosting for lin regression
Yury Babak created IGNITE-7150: -- Summary: Gradient boosting for lin regression Key: IGNITE-7150 URL: https://issues.apache.org/jira/browse/IGNITE-7150 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Currently for linreg we have only the analytical trainer (via QR decomposition). We want to add new trainer(implementation of Trainer interface) based on gradient boosting. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7096) Missed dependencies in examples pom files
Yury Babak created IGNITE-7096: -- Summary: Missed dependencies in examples pom files Key: IGNITE-7096 URL: https://issues.apache.org/jira/browse/IGNITE-7096 Project: Ignite Issue Type: Bug Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 Missed dependencies in ML profiles in pom-standalone.xml and pom-standalone-lgpl.xml. missed: commons-cli commons-cli 1.2 -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7174) Local MLP
Yury Babak created IGNITE-7174: -- Summary: Local MLP Key: IGNITE-7174 URL: https://issues.apache.org/jira/browse/IGNITE-7174 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Yury Babak Fix For: 2.4 local version of MLP -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7169) Missed javadoc for IgniteToDoubleFunction
Yury Babak created IGNITE-7169: -- Summary: Missed javadoc for IgniteToDoubleFunction Key: IGNITE-7169 URL: https://issues.apache.org/jira/browse/IGNITE-7169 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 Missed javadoc for IgniteToDoubleFunction -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-8741) [ML] Make a tutorial for data preprocessing
Yury Babak created IGNITE-8741: -- Summary: [ML] Make a tutorial for data preprocessing Key: IGNITE-8741 URL: https://issues.apache.org/jira/browse/IGNITE-8741 Project: Ignite Issue Type: Wish Components: ml Reporter: Yury Babak Assignee: Aleksey Zinoviev Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8795) Add ability to start and maintain TensorFlow cluster on top of Apache Ignite
Yury Babak created IGNITE-8795: -- Summary: Add ability to start and maintain TensorFlow cluster on top of Apache Ignite Key: IGNITE-8795 URL: https://issues.apache.org/jira/browse/IGNITE-8795 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Anton Dmitriev Fix For: 2.6 As described in the design document (https://docs.google.com/document/d/1jROIahK1rc7bSgOvhJhfpMqIGvht_IE8zn5NAt6x8ks/edit?usp=sharing) Distributed TensorFlow is based on TensorFlow cluster concept. It's a set of TensorFlow processes started among the cluster and available througth the gRPC interfaces. It's assumed that these processes contain heavy operations that requires data to be stored locally on the nodes where the processes running. Apache Ignite admits the data to be moved from one node to another as result of node failure of rebalancing. As result the TensorFlow cluster should be changed dynamically as well as TensorFlow Cache (follow-the-data strategy). -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8792) Introduction of TensorFlow integration module
Yury Babak created IGNITE-8792: -- Summary: Introduction of TensorFlow integration module Key: IGNITE-8792 URL: https://issues.apache.org/jira/browse/IGNITE-8792 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8680) Encoding categorical features with OneHotEncoder
Yury Babak created IGNITE-8680: -- Summary: Encoding categorical features with OneHotEncoder Key: IGNITE-8680 URL: https://issues.apache.org/jira/browse/IGNITE-8680 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Aleksey Zinoviev Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8679) Integration with tensorflow datasets
Yury Babak created IGNITE-8679: -- Summary: Integration with tensorflow datasets Key: IGNITE-8679 URL: https://issues.apache.org/jira/browse/IGNITE-8679 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Artem Malykh Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8670) Umbrella: TensorFlow integration
Yury Babak created IGNITE-8670: -- Summary: Umbrella: TensorFlow integration Key: IGNITE-8670 URL: https://issues.apache.org/jira/browse/IGNITE-8670 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8663) L1,L2 normalization
Yury Babak created IGNITE-8663: -- Summary: L1,L2 normalization Key: IGNITE-8663 URL: https://issues.apache.org/jira/browse/IGNITE-8663 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 We want to add L1 and L2 normalization using Model/Trainer API. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8664) Encoding categorical features
Yury Babak created IGNITE-8664: -- Summary: Encoding categorical features Key: IGNITE-8664 URL: https://issues.apache.org/jira/browse/IGNITE-8664 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8665) Umbrella: ML model validation for 2.6 release
Yury Babak created IGNITE-8665: -- Summary: Umbrella: ML model validation for 2.6 release Key: IGNITE-8665 URL: https://issues.apache.org/jira/browse/IGNITE-8665 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8666) Add ability of filtering data during datasets creation
Yury Babak created IGNITE-8666: -- Summary: Add ability of filtering data during datasets creation Key: IGNITE-8666 URL: https://issues.apache.org/jira/browse/IGNITE-8666 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8668) K-fold cross validation of models
Yury Babak created IGNITE-8668: -- Summary: K-fold cross validation of models Key: IGNITE-8668 URL: https://issues.apache.org/jira/browse/IGNITE-8668 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8662) Umbrella: ML data preprocessing for 2.6 release
Yury Babak created IGNITE-8662: -- Summary: Umbrella: ML data preprocessing for 2.6 release Key: IGNITE-8662 URL: https://issues.apache.org/jira/browse/IGNITE-8662 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8667) Splitting of dataset to test and training sets
Yury Babak created IGNITE-8667: -- Summary: Splitting of dataset to test and training sets Key: IGNITE-8667 URL: https://issues.apache.org/jira/browse/IGNITE-8667 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Anton Dmitriev Fix For: 2.6 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8669) Model estimation
Yury Babak created IGNITE-8669: -- Summary: Model estimation Key: IGNITE-8669 URL: https://issues.apache.org/jira/browse/IGNITE-8669 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 We want to have the common mechanism for model estimation. For estimation we want to have: * Accuracy/precision/recall * F score * TPR/FRP * ROC AUC -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8907) [ML] Using vectors in featureExtractor
Yury Babak created IGNITE-8907: -- Summary: [ML] Using vectors in featureExtractor Key: IGNITE-8907 URL: https://issues.apache.org/jira/browse/IGNITE-8907 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Alexey Platonov Fix For: 2.7 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-8867) Bootstrapping for learning sample
Yury Babak created IGNITE-8867: -- Summary: Bootstrapping for learning sample Key: IGNITE-8867 URL: https://issues.apache.org/jira/browse/IGNITE-8867 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.7 Need to implement bootstrapping algorithm in Bagging-classifier -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7297) Javadoc warning for RProp in MLP
Yury Babak created IGNITE-7297: -- Summary: Javadoc warning for RProp in MLP Key: IGNITE-7297 URL: https://issues.apache.org/jira/browse/IGNITE-7297 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.4 [Step 7/7] [WARNING] /data/teamcity/work/bd85361428dcdb1/modules/ml/src/main/java/org/apache/ignite/ml/nn/updaters/RPropUpdater.java:32: warning - Tag @see: missing final '>': "https://paginas.fe.up.pt/~ee02162/dissertacao/RPROP%20paper.pdf;>https://paginas.fe.up.pt/~ee02162/dissertacao/RPROP%20paper.pdf." [20:56:44][Step 7/7] [WARNING] /data/teamcity/work/bd85361428dcdb1/modules/ml/src/main/java/org/apache/ignite/ml/nn/updaters/RPropUpdaterParams.java:28: warning - Tag @see: missing final '>': "https://paginas.fe.up.pt/~ee02162/dissertacao/RPROP%20paper.pdf;>https://paginas.fe.up.pt/~ee02162/dissertacao/RPROP%20paper.pdf." -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7438) LSQR: Sparse Equations and Least Squares for Lin Regression
Yury Babak created IGNITE-7438: -- Summary: LSQR: Sparse Equations and Least Squares for Lin Regression Key: IGNITE-7438 URL: https://issues.apache.org/jira/browse/IGNITE-7438 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Anton Dmitriev We to implemet LSQR trainer for lin regresstion. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7437) Partition based dataset implementation
Yury Babak created IGNITE-7437: -- Summary: Partition based dataset implementation Key: IGNITE-7437 URL: https://issues.apache.org/jira/browse/IGNITE-7437 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Anton Dmitriev We want to implement our dataset based on entire partition instead of key sets. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7484) Documentation for new lin reg trainer.
Yury Babak created IGNITE-7484: -- Summary: Documentation for new lin reg trainer. Key: IGNITE-7484 URL: https://issues.apache.org/jira/browse/IGNITE-7484 Project: Ignite Issue Type: Task Components: ml Reporter: Yury Babak -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7503) MLP documentation
Yury Babak created IGNITE-7503: -- Summary: MLP documentation Key: IGNITE-7503 URL: https://issues.apache.org/jira/browse/IGNITE-7503 Project: Ignite Issue Type: Sub-task Components: documentation, ml Reporter: Yury Babak Assignee: Artem Malykh Fix For: 2.4 A need to add documentation about MLP -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7504) Decision tree documentation
Yury Babak created IGNITE-7504: -- Summary: Decision tree documentation Key: IGNITE-7504 URL: https://issues.apache.org/jira/browse/IGNITE-7504 Project: Ignite Issue Type: Task Components: documentation, ml Reporter: Yury Babak Assignee: Artem Malykh Fix For: 2.4 We want to add Decision tree documentation -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7643) Broken javadoc in partitioned dataset
Yury Babak created IGNITE-7643: -- Summary: Broken javadoc in partitioned dataset Key: IGNITE-7643 URL: https://issues.apache.org/jira/browse/IGNITE-7643 Project: Ignite Issue Type: Task Components: ml Affects Versions: 2.5 Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.5 [22:25:12][Step 7/7] [WARNING] Javadoc Warnings [22:25:12][Step 7/7] [WARNING] /data/teamcity/work/bd85361428dcdb1/examples/src/main/java/org/apache/ignite/examples/ml/dataset/AlgorithmSpecificDatasetExample.java:51: warning - Tag @link: reference not found: AlgorithmSpecificDataset [22:25:12][Step 7/7] [WARNING] /data/teamcity/work/bd85361428dcdb1/examples/src/main/java/org/apache/ignite/examples/ml/dataset/AlgorithmSpecificDatasetExample.java:51: warning - Tag @link: reference not found: AlgorithmSpecificPartitionContext -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7716) Red selftest in ML examples
Yury Babak created IGNITE-7716: -- Summary: Red selftest in ML examples Key: IGNITE-7716 URL: https://issues.apache.org/jira/browse/IGNITE-7716 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Fix For: 2.5 https://ci.ignite.apache.org/project.html?tab=testDetails=IgniteTests24Java8=1447870893775475761 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7741) Fix javadoc for QR factorization
Yury Babak created IGNITE-7741: -- Summary: Fix javadoc for QR factorization Key: IGNITE-7741 URL: https://issues.apache.org/jira/browse/IGNITE-7741 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Wrong javadoc for QR factorization. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-7332) Create selftest suite for ml examples
Yury Babak created IGNITE-7332: -- Summary: Create selftest suite for ml examples Key: IGNITE-7332 URL: https://issues.apache.org/jira/browse/IGNITE-7332 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.5 We want to add self test suite for our examples like we have for java8 examples. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-7350) Distributed MLP cleanup/refactoring
Yury Babak created IGNITE-7350: -- Summary: Distributed MLP cleanup/refactoring Key: IGNITE-7350 URL: https://issues.apache.org/jira/browse/IGNITE-7350 Project: Ignite Issue Type: Bug Components: ml Reporter: Yury Babak Assignee: Yury Babak Current state of MLP not so good, so we need improve it and may be rewrite some parts of this NN. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Created] (IGNITE-8840) Random Forest
Yury Babak created IGNITE-8840: -- Summary: Random Forest Key: IGNITE-8840 URL: https://issues.apache.org/jira/browse/IGNITE-8840 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Fix For: 2.6 We want to implement random forest algorithm. It should be based on our implementation of decision trees. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9158) [ML] Pipeline
Yury Babak created IGNITE-9158: -- Summary: [ML] Pipeline Key: IGNITE-9158 URL: https://issues.apache.org/jira/browse/IGNITE-9158 Project: Ignite Issue Type: New Feature Components: ml Reporter: Yury Babak Assignee: Aleksey Zinoviev Fix For: 2.7 We want to implement our own pipeline for ML operations. More details in [dev-list|http://apache-ignite-developers.2346864.n4.nabble.com/ML-Machine-Learning-Pipeline-Improvement-tt32772.html] -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9237) [ML] Random forest optimization
Yury Babak created IGNITE-9237: -- Summary: [ML] Random forest optimization Key: IGNITE-9237 URL: https://issues.apache.org/jira/browse/IGNITE-9237 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Alexey Platonov Fix For: 2.7 We need to implement best split selection by statistics over impurity data and share this data for several nodes in several trees while learning process. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9387) [ML] Model updating
Yury Babak created IGNITE-9387: -- Summary: [ML] Model updating Key: IGNITE-9387 URL: https://issues.apache.org/jira/browse/IGNITE-9387 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Alexey Platonov Fix For: 2.7 In trainer interface we need to support model updating by batches after first training -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9415) [ML] Using sparce vectors in LSQR and MLP
Yury Babak created IGNITE-9415: -- Summary: [ML] Using sparce vectors in LSQR and MLP Key: IGNITE-9415 URL: https://issues.apache.org/jira/browse/IGNITE-9415 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.7 We need to investigate and apply sparce vectors support in BLAS for LSQR and MLP (or implement own version) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9414) [ML] Using sparce vectors in Tree-based algorithms.
Yury Babak created IGNITE-9414: -- Summary: [ML] Using sparce vectors in Tree-based algorithms. Key: IGNITE-9414 URL: https://issues.apache.org/jira/browse/IGNITE-9414 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.7 We need to support sparce vectors in DecisionTrees, RF, GDB -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9412) [ML] GDB convergence by error support.
Yury Babak created IGNITE-9412: -- Summary: [ML] GDB convergence by error support. Key: IGNITE-9412 URL: https://issues.apache.org/jira/browse/IGNITE-9412 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.7 We need to support early training interruption when GDB has small error rate on learning sample -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9413) [ML] Learning rate optimization for GDB.
Yury Babak created IGNITE-9413: -- Summary: [ML] Learning rate optimization for GDB. Key: IGNITE-9413 URL: https://issues.apache.org/jira/browse/IGNITE-9413 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Fix For: 2.7 We need to support learning rate optimization while training for MSE-loss and Log-loss -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9034) [ML] Add Estimator API support to TensorFlow cluster on top of Apache Ignite
Yury Babak created IGNITE-9034: -- Summary: [ML] Add Estimator API support to TensorFlow cluster on top of Apache Ignite Key: IGNITE-9034 URL: https://issues.apache.org/jira/browse/IGNITE-9034 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Anton Dmitriev Fix For: 2.7 -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9064) Decision tree optimization
Yury Babak created IGNITE-9064: -- Summary: Decision tree optimization Key: IGNITE-9064 URL: https://issues.apache.org/jira/browse/IGNITE-9064 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Alexey Platonov Fix For: 2.7 We need to optimize impurity function calculation by additional index structure for all sorted features and reusing it in learning iterations. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9065) Gradient boosting optimization
Yury Babak created IGNITE-9065: -- Summary: Gradient boosting optimization Key: IGNITE-9065 URL: https://issues.apache.org/jira/browse/IGNITE-9065 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Alexey Platonov Fix For: 2.7 We need to optimize GDB learning by reusing same index for learning decision trees. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9021) [ML] Refactor vectors to dence/sparse
Yury Babak created IGNITE-9021: -- Summary: [ML] Refactor vectors to dence/sparse Key: IGNITE-9021 URL: https://issues.apache.org/jira/browse/IGNITE-9021 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Aleksey Zinoviev Fix For: 2.7 We want to remove all unused implementations of Vector interface. Same for matrices. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9437) [ML] Add performance benchmarks
Yury Babak created IGNITE-9437: -- Summary: [ML] Add performance benchmarks Key: IGNITE-9437 URL: https://issues.apache.org/jira/browse/IGNITE-9437 Project: Ignite Issue Type: Improvement Components: ml Reporter: Yury Babak Assignee: Oleg Ignatenko Fix For: 2.7 We want to have some performance benchmarks for ML algorithms -- This message was sent by Atlassian JIRA (v7.6.3#76005)