Contact

Claudia Beleites
Chemometric Consulting
Södeler Weg 19
61200 Wölfersheim/Germany
e-mail: claudia dot beleites at chemometrix dot eu

All Posts

Excerpts

New contact data and email, and a bit of maintenance

Contact address as well as emails were updated to my new freelancing email and address. Imports were updated as well.

None of these updates affects calculations or other user visible behaviour.

... read more ...

Slight overhaul to keep up with R and roxygen2

Some updates were necessary because of new requirements for R’s package infrastructure. None of these updates affects calculations or other user visible behaviour.

... read more ...

Citation data corrected

Fixed the Bibtex citation information.

... read more ...

softclassval Version 1.0

With the theory paper being printed, softclassval is now mature to gain Version number 1.0.

... read more ...

checkrp exported

Function checkrp which does consistency checks and possibly recycles the reference to fit to the predictions is now exported. This saves time if multiple performance measures are to be calculated on the same predictions.

... read more ...

Validation of Soft Classification Models using Partial Class Memberships: An Extended Concept of Sensitivity & Co. applied to Grading of Astrocytoma Tissues

Paper deriving and explaining the theory behind softclassval

We use partial class memberships in soft classification to model uncertain labelling and mixtures of classes. Partial class memberships are not restricted to predictions, but may also occur in reference labels (ground truth, gold standard diagnosis) for training and validation data.

... read more ...

Unit Tests

One of the reviewers of the theory paper asks how softclassval ensures computational correctness.

Unit Tests

softclassval uses svUnit for unit testing.

... read more ...

Validation of Soft Classifiers for Cells and Tissues

Medical diagnosis of cells and tissues is an important aim in biospectroscopy. The data analytical task involved frequently is classification. Classification traditionally assumes both reference and prediction to be hard, i.e. stating exactly one of the defined classes. Like fuzzy cluster analysis, soft classification uses partial memberships rather than hard labels, thus expressing uncertainty or mixed cell populations.

... read more ...

Spectroscopic Data in R and Validation of Soft Classifiers: Classifying Cells and Tissues by Raman Spectroscopy

Medical diagnosis of cells and tissues is an important aim in biospectroscopy. The data analytical task involved frequently is classification. Classification traditionally assumes both reference and prediction to be hard, i.e. stating exactly one of the defined classes. In reality, the reference diagnoses may suffer from substantial uncertainty, or the sample can comprise a mixture of the underlying classes, e.g. if sample heterogeneity is not resolved or if the sample is actually undergoing a transition from one class to another (e.g. rather continuous de-differentiation of tumour tissues). Such samples may be labelled with partial or soft class memberships.

... read more ...

Raman spectroscopic grading of astrocytoma tissues: using soft reference information

Paper about the application that triggered development of softclassval

Gliomas are the most frequent primary brain tumours. During neurosurgical treatment, locating the exact tumour border is often difficult. This study assesses grading of astrocytomas based on Raman spectroscopy for a future application in intra-surgical guidance.

... read more ...

Soft Classification Models - Calculating Sensitivity and Specificity

... read more ...