Mahout 1.0 Features by Engine (unreleased)

All Single Machine and MapReduce features are available in the latest 0.9 release

Single Machine MapReduce Spark H2O Flink
Mahout Math-Scala Core Library and Scala DSL
Mahout Distributed BLAS. Distributed Row Matrix API with R and Matlab like operators. Distributed ALS, SPCA, SSVD, thin-QR. Similarity Analysis. x x in development
Mahout Interactive Shell
Interactive REPL shell for Spark optimized Mahout DSL x
Collaborative Filtering with CLI Drivers
User-Based Collaborative Filtering x x
Item-Based Collaborative Filtering x x x
Matrix Factorization with ALS x x
Matrix Factorization with ALS on Implicit Feedback x x
Weighted Matrix Factorization, SVD++ x
Classification with CLI Drivers
Logistic Regression - trained via SGD x
Naive Bayes / Complementary Naive Bayes x x in development
Random Forest x
Hidden Markov Models x
Multilayer Perceptron x
Clustering with CLI Drivers
Canopy Clustering deprecated deprecated
k-Means Clustering x x
Fuzzy k-Means x x
Streaming k-Means x x
Spectral Clustering x
Dimensionality Reduction with CLI Drivers - note: most scala-based dimensionality reduction algorithms are available through the Mahout Math-Scala Core Library for all engines
Singular Value Decomposition x x
Lanczos Algorithm x x
Stochastic SVD x x
PCA (via Stochastic SVD) x x
QR Decomposition x x
Topic Models
Latent Dirichlet Allocation x x
Miscellaneous
RowSimilarityJob x x
ConcatMatrices x
Collocations x
Sparse TF-IDF Vectors from Text x
XML Parsing x
Email Archive Parsing x
Lucene Integration x
Evolutionary Processes x