Open source · Research · Education

Research methods made practical

I develop and maintain open-source software that turns research in data mining, machine learning, and optimization into reusable tools. Most projects are R packages, with selected Python interfaces, Bioconductor packages, and teaching resources.

Platforms: R-universe · GitHub · PyPI · Bioconductor

Pattern mining

Discover, represent, visualize, and use frequent patterns, association rules, and sequences.

arules

R · Python interface

Core infrastructure and fast algorithms for mining and analyzing frequent itemsets and association rules.

arulespy

Python

A Python interface for exploring association rules and frequent itemsets with arules.

arulesViz

R

Static and interactive visualization techniques for association rules and itemsets.

arulesSequences

R

Mining and manipulating frequent sequences using the SPADE algorithm family.

arulesCBA

R

Associative classification with CBA, CMAR, CPAR, FOIL, PART, PRM, RCAR, and RIPPER.

arulesNBMiner

R

Model-based mining of NB-frequent itemsets and NB-precise association rules.

Clustering and data streams

Analyze density-based structure and evolving data with efficient, extensible algorithms.

dbscan

R

Fast implementations of DBSCAN, HDBSCAN, OPTICS, LOF, and nearest-neighbor search.

stream

R

An extensible framework for data-stream clustering, evaluation, and experimentation.

streamConnect

R

Connect stream-mining components using sockets and deploy them as web services.

streamMOA

R · Java

An interface to clustering algorithms from the MOA data-stream mining framework.

rEMM

R

Extensible Markov Models that add temporal structure to data-stream clustering.

Decision-making and optimization

Model sequential decisions and solve ordering, assignment, routing, and control problems.

pomdp

R

Define, solve, simulate, and analyze partially observable Markov decision processes.

pomdpSolve

R · C

A convenient R distribution and interface for Anthony Cassandra's pomdp-solve software.

TSP

R

Infrastructure and algorithms for traveling salesperson problems, including a Concorde interface.

seriation

R

Reorder matrices, dissimilarities, and dendrograms to reveal structure and improve visualization.

qap

R

Heuristics for the quadratic assignment problem, including simulated annealing.

Recommender systems

Build, evaluate, and compare recommendation algorithms using a shared framework.

recommenderlab

R

Research and evaluation infrastructure for collaborative filtering and other recommender algorithms.

Bioinformatics

Connect established biological-sequence tools with reproducible R and Bioconductor workflows.

rBLAST

R · Bioconductor

An R interface to BLAST for searching genetic sequence databases within Bioconductor workflows.

Educational software and materials

Open examples, assignments, and small tools for learning data mining, artificial intelligence, reinforcement learning, and programming.

Data Mining with R

R · Book

Documented R examples accompanying Introduction to Data Mining by Tan, Steinbach, Karpatne, and Kumar.

Artificial Intelligence

Python · Course

Lecture material, compact code examples, and assignments based on Artificial Intelligence: A Modern Approach.

Reinforcement Learning

Python · Course

Teaching material following Sutton and Barto's Reinforcement Learning: An Introduction.

gym-classics2

Python

Classic discrete finite Markov decision processes for reinforcement-learning experiments and instruction.

Small teaching tools

R · C++

Distribution-fitting examples and Gridhunt2, a small game for teaching object-oriented programming concepts.

Archived projects

These projects remain available for reproducibility and historical use but are no longer the recommended starting point for new work.

rRDP

R · Bioconductor

An R interface to the Ribosomal Database Project classifier for classifying ribosomal RNA sequences.

rMSA

R

An R interface to multiple-sequence alignment tools including ClustalW, MAFFT, MUSCLE, and Kalign.

QuasiAlign

R

Prototype software for efficient biological-sequence analysis using alignment-free methods.