Papers by Smaranda Belciug
Encyclopedia of Biomedical Engineering
Artificial Intelligence in Cancer
The aim of this chapter is to explore the untapped field of artificial intelligence applied in ra... more The aim of this chapter is to explore the untapped field of artificial intelligence applied in radiotherapy. We begin with reviewing different types of treatment, followed by a little bit of history. Later on the presentation is moved onto new AI systems that can speed up the workflow and ease the job of radiation oncologists. The chapter ends with discussions regarding ethical and moral aspects of applying artificial intelligence in radiation therapy. Is artificial intelligence snake oil? Could artificial intelligence end up being racist? The answers to these questions lie in the pages of this chapter.
The Intelligent Decision Support Systems (IDSSs) represent an interdisciplinary research domain b... more The Intelligent Decision Support Systems (IDSSs) represent an interdisciplinary research domain bringing together Artificial Intelligence/Machine Learning (AI/ML), Decision Science (DS), and Information Systems (IS). IDSS refers to the use of AI/ML techniques in decision support systems. In this context, it should be emphasized the special role of statistical learning (SL) in the process of training algorithms from data. The purpose of this chapter is to provide a short review of some of the state-of-the-art AI/ML algorithms, seen as intelligent tools used in the medical decision-making, along with some important applications in the automated medical diagnosis of some major chronic diseases (MCDs). In addition, we aim to present an interesting approach to develop novel IDSS inspired by the evolutionary paradigm.
Artificial Intelligence in Cancer
The aim of this chapter is to discuss what happens after the cancer patient enters the remission ... more The aim of this chapter is to discuss what happens after the cancer patient enters the remission phase. We discuss types of remission, see the wonders of the spontaneous ones, and learn to compute the odds of recurrence using AI methods. New AI methods are presented together with their application in the field. Some of the scholarly articles presented in this chapter use algorithms and statistical tests that have been explained in former chapters.
Annals of the University of Craiova - Mathematics and Computer Science Series, 2018
Extreme Learning Machine (ELM) is a single-hidden layer feedforward neural network, where the wei... more Extreme Learning Machine (ELM) is a single-hidden layer feedforward neural network, where the weights between the input and hidden layer are randomly generated and never updated, whereas the hidden-output weights are analytically computed. Theoretical studies have shown that ELM maintains the universal approximation capability. Artificial Intelligence applied in automated medical diagnosis is problematic due to the high risk of overfitting the data, because of the large number of attributes. The goal of this paper is to propose a feature selection (FS) mechanism based on Ant Colony Optimization (ACO), in order to speed up the computational process of the ELM. The proposed model has been tested on three publicly available high-dimensional datasets.
The aim of this chapter is to show how Artificial Intelligence can be applied in oncology. First ... more The aim of this chapter is to show how Artificial Intelligence can be applied in oncology. First we shall see what a medical oncologist does, what type of chemotherapies are employed, followed by how could Artificial Intelligence help. We shall go through finding connections between chemotherapy drugs and immune systems, what is the impact of hormonal therapy versus chemotherapy in breast cancer, immunotherapy and lots of other interesting facts.
Neural networks are well-known for their effectiveness as classifiers. In this context, feature s... more Neural networks are well-known for their effectiveness as classifiers. In this context, feature selection has become a usefulness technique to make the classifier faster, cost-effective, and more accurate. The effect of a random forests-based feature selection on the classification accuracy of a multi-layer perceptron has been explored in this paper. Up to 6% improvement in classification accuracy and 40% improvement in computation speed has been observed when using the tandem model on five real-world publically available medical datasets regarding colon cancer, breast cancer, diabetes, thyroid and fetal heartbeat databases.
Studies in Informatics and Control
This paper proposes the application to the liver fibrosis stadialization of a novel training tech... more This paper proposes the application to the liver fibrosis stadialization of a novel training technique of feed-forward neural networks based on the Bayesian paradigm. Using the Pearson’s r correlation coefficient instead of the standard backpropagation algorithm to update the synaptic weights of a multi-layer perceptron, the proposed model is compared with traditional machine learning algorithms (standard MLP, RBF, PNN, SVM) using a real-life liver fibrosis dataset. The statistical comparison results indicated that the Bayesian-trained MLP proved to be at least as efficient as its classic competitors.
Annals of the University of Craiova - Mathematics and Computer Science Series, 2008
In many domains, a range of input variables are considered, not clearly which of them are most us... more In many domains, a range of input variables are considered, not clearly which of them are most useful, or indeed are needed at all. Data are often collected on variables that are not only correlated, but also are large in number. This makes the data process, interpretation and detection of its structure difficult. Feature selection is a pattern recognition approach to choose important variables according to some criteria, in order to improve the decision process by removing the redundant information. The intent of this work is to provide a feature selection approach, based on the analysis of correlations between the explanatory (input) variables and the outcome variables, to improve the classification process of the liver fibrosis stages, using both the naive Bayes classifier and the probabilistic neural network model. 2000 Mathematics Subject Classification. 62C10; 92B20.
Studies in Informatics and Control, 2021
Annals of the University of Craiova - Mathematics and Computer Science Series, 2009
Patient management is a very challenging problem of the health care system. Grouping patients acc... more Patient management is a very challenging problem of the health care system. Grouping patients according to their Length of Stay (LoS) in the hospital leads to a better planning of bed allocation, and patient admission and discharge. The aim of this paper is to statistically evaluate and cluster the data using the agglomerative hierarchical clustering
Annals of the University of Craiova - Mathematics and Computer Science Series, 2019
Fairly recently, extreme learning machine (ELM) has been proposed as a single-hidden layer feedfo... more Fairly recently, extreme learning machine (ELM) has been proposed as a single-hidden layer feedforward neural network (SLFN), where the input weights are randomly initiated and never updated, and the output weights are analytically computed. Setting the parameters of the hidden layer randomly may not be always effective if the function that is learned is not simple and the amount of labeled data is not small, even if theoretical studies have shown that ELM maintains the universal approximation capability. To address this issue, we propose a new approach inspired by the Bayesian paradigm as an alternative to the random initiation of the hidden node parameters. The idea behind this model is that we can use the information (prior knowledge) about a certain labeled data through the correlation between attributes and decision classes. The prior knowledge is acquired through the Goodman-Kruskal Gamma rank correlation between attributes and labels, assuming that the input weights should be...
Learning and Analytics in Intelligent Systems
Slowly, but steadily, we have reached the last chapter of our journey. Each part of the book brou... more Slowly, but steadily, we have reached the last chapter of our journey. Each part of the book brought a piece of puzzle and placed it in the right position. At last, the compass will guide us to our ultimate goal: intelligent digital healthcare revolution!
Handbook of Artificial Intelligence in Healthcare
Annals of the University of Craiova - Mathematics and Computer Science Series, 2017
Predicting stock market price over time is an important issue that needs to be addressed. The clo... more Predicting stock market price over time is an important issue that needs to be addressed. The closing price for the stocks is strongly correlated to many variables, so, in order to make the prediction more accurate one must try to detect the most important ones. The main objective of this research is to predict the on day closing price of ten companies enlisted in the Romanian Stock Market using different machine learning (ML) techniques, such as multiple linear regression (MR), and multilayer perceptron neural networks (MLP) setup for regression.
Uploads
Papers by Smaranda Belciug