University homepage Suomenkielinen versio puuttuu Inte på svenska In english
University of Helsinki Department of Computer Science
 

Department of Computer Science

Software

This page contains some software I have written. The programs are provided as they are, without any guarantees. Use at your own risk!

Cryo-electron microscopy

Orient Express
A software system for determining relative orientations of cryo-electron images taken from random directions of identical objects using common lines techinques.

Related papers
Mielikäinen, Ravantti, Ukkonen: The Computational complexity of orientation search in cryo-electron microscopy, ICCS 2004
Mielikäinen: Läpivalaisukuvien orientointi, MSc Thesis.

Sinogram denoiser
A program for denoising cryo-electron microscope images using sinograms.

Related papers
Mielikäinen, Ravantti: Sinogram denoising of cryo-electron microscopy images, ICCSA 2005
Ravantti: Computational methods for reconstructing macromolecular complexes from cryo-electron microscopy images, PhD Thesis

Pattern discovery

Intersector
A C++ implementation of a method for mining closed itemsets with various constraints based on the definition of closed itemsets as intersections of transactions.

Related papers
Mielikäinen: Intersecting Data to Closed Sets with Constraints, FIMI 2003

Itemset collection condensation by discretization
Perl scripts for condensing frequent itemset collections by discretization.

Related papers
Mielikäinen: Frequency-based views to pattern collections, Discrete Applied Mathematics 154(7):1113-1139
Mielikäinen: Summarization techniques for pattern collections in data mining, PhD Thesis

Closed itemset approximations by pattern ordering
Perl scripts for approximating closed itemsets by applying the greedy pattern ordering with frequency estimates being maximums of known supersets' frequencies.

Related papers
Mielikäinen, Mannila: The pattern ordering problem, PKDD 2003
Mielikäinen: Summarization techniques for pattern collections in data mining, PhD Thesis

Chaining patterns
C programs for minimum and minimal chain partitions of itemset collections using maximum and maximal matching, respectively.

Related papers
Mielikäinen: Chaining patterns, Discovery Science 2003
Mielikäinen: Summarization techniques for pattern collections in data mining, PhD Thesis

Deanonymization of anonymized transaction databases
Perl scripts constructing mixed integer linear programs in lp_solve format to analyze how well the item names in an itemset collection can be matched to items of anonymized transaction databases.

Related papers
Mielikäinen: Privacy problems with anonymized transaction databases, Discovery Science 2004

Mining chains of relations
Programs for answering data mining queries in a three-level graph. Some of the queries use a mixed integer linear program solver such as lp_solve, and some need a closed frequent itemset miner available from FIMI repository.

Related papers
Afrati, Das, Gionis, Mannila, Mielikäinen, Tsaparas: Mining chains of relations, ICDM 2005

Swap randomization for assessing data mining results
Tools for assesing data mining results from binary data by comparing the results to the results on random datasets with the same row and column margins. (I wrote the perl programs. Other implementations in the directory are due to Aristides Gionis and Panayiotis Tsaparas.)

Related papers
Gionis, Mannila, Mielikäinen, Tsaparas: Assessing data mining results via swap randomization, KDD 2006

Itemset Support Query Anwering
Algorithms for querying itemset supports using trie representations of various condensed representations of frequent itemsets.

Related papers
Mielikäinen, Panov, Dzeroski: Itemset Support Queries using Frequent Itemsets and Their Condensed Representations, Discovery Science 2006