Papers by Marina Velikova
Lecture Notes in Computer Science, 2015
Conscious Cogn, 2006
This dissertation studies the incorporation of monotonicity constraints as a type of domain knowl... more This dissertation studies the incorporation of monotonicity constraints as a type of domain knowledge into a data mining process. Monotonicity constraints are enforced at two stages¿data preparation and data modeling. The main contributions of the research are a novel procedure to test the degree of monotonicity of a real data set, a greedy algorithm to transform non-monotone into monotone data, and extended and novel approaches for building monotone decision models. The results from simulation and real case studies show that enforcing monotonicity can considerably improve knowledge discovery and facilitate the decision-making process for end-users by deriving more accurate, stable and plausible decision models.
Proceedings of the Twenty Third International Joint Conference on Artificial Intelligence, 2013
ABSTRACT Probabilistic logics combine the expressive power of logic with the ability to reason wi... more ABSTRACT Probabilistic logics combine the expressive power of logic with the ability to reason with uncertainty. Several probabilistic logic languages have been proposed in the past, each of them with their own features. In this paper, we propose a new probabilistic constraint logic programming language, which combines constraint logic programming with probabilistic reasoning. The language supports modeling of discrete as well as continuous probability distributions by expressing constraints on random variables. We introduce the declarative semantics of this language, present an exact inference algorithm to derive bounds on the joint probability distributions consistent with the specified constraints, and give experimental results. The results obtained are encouraging, indicating that inference in our language is feasible for solving challenging problems.
Screening mammography is of a crucial value for the early detection of breast cancer. Computer-ai... more Screening mammography is of a crucial value for the early detection of breast cancer. Computer-aided (CAD) systems are developed in order to given support to radiologists on the detection of breast lesions and correct interpretation of cancers. As images of each breast are usually taken from different projections, yielding multiple views of the same breast, accurate detection of abnormal findings requires comparison between those views. Usually CAD systems do not take this information into account, assuming that mammographic views are independent. We, on the other hand, introduce in this paper different aspects of breast cancer modelling by combining multiple mammographic views using Bayesian network theory. The models we propose are based on advanced raw image data analysis and domain knowledge of mammographic screening and have the aim to evaluate alternatives for improving lesion detection accuracy.
Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2011
Lecture Notes in Computer Science, 2011
Procedia Computer Science, 2015
Deliver "actionable" intelligence instead of just raw information -this is what the Metis researc... more Deliver "actionable" intelligence instead of just raw information -this is what the Metis research project pursues for supporting operational work in domains characterized by constantly evolving situations with a diversity of entities, complex interactions and high-level uncertainty in the information gathered. Operating effectively in such domains requires robust, on-the-fly and context-based information reasoning, which goes beyond human capabilities. In this paper a real-time reference architecture is presented employing and integrating several state-of-the-art computing technologies for automated and consolidated 'situational understanding'. In particular, outlined are the innovative components (i) for fusing of and reasoning on uncertain information based on probabilistic logic and (ii) for a complementary interactive visualization disclosing the system's line of reasoning inferred from the domain model and provided evidence. The architecture has been realized as a fully demonstrable proof of concept and its applied value is illustrated in a number of real and fictive cases from the domain of maritime safety and security.
Lecture Notes in Computer Science, 2012
ABSTRACT In various domains, such as security and surveillance, a large amount of information fro... more ABSTRACT In various domains, such as security and surveillance, a large amount of information from heterogeneous sources is continuously gathered to identify and prevent potential threats, but it is unknown in advance what the observed entity of interest should look like. The quality of the decisions made depends, of course, on the quality of the information they are based on. In this paper, we propose a novel method for assessing the quality of information taking into account uncertainty. Two properties --- soundness and completeness --- of the information are used to define the notion of information quality and their expected values are defined using a probabilistic model output. Simulation experiments with data from a maritime scenario demonstrates the usage of the proposed method and its potential for decision support in complex tasks such as surveillance.
Healthcare Technology Letters, 2014
In the context of home-based healthcare monitoring systems, it is desirable that the results obta... more In the context of home-based healthcare monitoring systems, it is desirable that the results obtained from biochemical tests - tests of various body fluids such as blood and urine - are objective and automatically generated to reduce the number of man-made errors. The authors present the StripTest reader - an innovative smartphone-based interpreter of biochemical tests based on paper-based strip colour using image processing techniques. The working principles of the reader include image acquisition of the colour strip pads using the camera phone, analysing the images within the phone and comparing them with reference colours provided by the manufacturer to obtain the test result. The detection of kidney damage was used as a scenario to illustrate the application of, and test, the StripTest reader. An extensive evaluation using laboratory and human urine samples demonstrates the reader's accuracy and precision of detection, indicating the successful development of a cheap, mobile and smart reader for home-monitoring of kidney functioning, which can facilitate the early detection of health problems and a timely treatment intervention.
2012 25th IEEE International Symposium on Computer-Based Medical Systems (CBMS), 2012
ABSTRACT In this paper we present innovative research for the automatic interpretation of biochem... more ABSTRACT In this paper we present innovative research for the automatic interpretation of biochemical test strip color by a smartphone using image processing techniques. Urinalysis is the current application for these techniques. Our mobile application captures images of the color pads on strips using the camera phone, then analyzes automatically the images within the device itself and compares these against reference color pads to obtain a final classification. As a test scenario we focus on the detection of proteinuria, i.e., leakage of protein into the urine, for the detection of which strips for protein and creatinine are used. We performed an initial laboratory evaluation using specially prepared concentrations to check the accuracy and precision of detection. The results obtained are encouraging and show that the proposed technique has a good potential for the development of cheap, mobile and smart home-based readers for the early detection of health problems.
ABSTRACT Probabilistic logics combine the expressive power of logic with the ability to reason wi... more ABSTRACT Probabilistic logics combine the expressive power of logic with the ability to reason with uncertainty. Several probabilistic logic languages have been proposed in the past, each of them with their own features. In this paper, we propose a new probabilistic constraint logic programming language, which combines constraint logic programming with probabilistic reasoning. The language supports modeling of discrete as well as continuous probability distributions by expressing constraints on random variables. We introduce the declarative semantics of this language, present an exact inference algorithm to derive bounds on the joint probability distributions consistent with the specified constraints, and give experimental results. The results obtained are encouraging, indicating that inference in our language is feasible for solving challenging problems.
Uploads
Papers by Marina Velikova