TREE STRUCTURES OF LINEAR THRESHOLD UNITS FOR THE CLASSIFICATION OF NUMERIC EXAMPLES

A decision tree-based system for learning from numeric data is described. Results from linear algebra (pseudoinverse matrices) help the system to generate decision trees where the nodes are represented by linear threshold units minimizing the mean square error. The system's capability to provide good classifications with small decision trees is demonstrated on artificial and benchmark data.     

A rough set approach to reasoning under uncertainty

Abstract

Reasoning with uncertain information is a problem of key importance when dealing with information about the real world. Obtaining the precise numbers required by many uncertainty handling formalisms can be a problem. The theory of rough sets makes it possible to handle uncertainty without the need for precise numbers, and so has some advantages in such situations. This paper presents an introduction to various forms of reasoning under uncertainty that are based on rough sets. In particular, a number of sets of numerical and symbolic truth values which may be used to augment propositional logic are developed, and a semantics for these values is provided based upon the notion of possible worlds. Methods of combining the truth values are developed so that they may be propagated when augmented logic formulae are combined, and their use is demonstrated in theorem proving.     

Knockdown of apoptosis-inducing factor disrupts function of respiratory complex

Recent findings suggest that apoptotic protein apoptosis-inducing factor (AIF) may also play an important non-apoptotic function inside mitochondria. AIF was proposed to be an important component of respiratory chain complex I that is the major producer of superoxide radical. The possible role of AIF is still controversial. Superoxide production could be used as a valuable measure of complex I function, because the majority of superoxide is produced there. Therefore, we employed superoxide-specific mitochondrial fluorescence dye for detection of superoxide production. We studied an impact of AIF knockdown on function of mitochondrial complex I by analyzing superoxide production in selected cell lines. Our results show that tumoral telomerase-positive (TP) AIF knockdown cell lines display significant increase in superoxide production in comparison to control cells, while a non-tumoral cell line and tumoral telomerase-negative cell lines with alternative lengthening of telomeres (ALT) show a decrease in superoxide production. According to these results, we can conclude that AIF knockdown disrupts function of complex I and therefore increases the superoxide production in mitochondria. The distinct effect of AIF depletion in various cell lines could result from recently discovered activity of telomerase in mitochondria of TP cancer cells, but this hypothesis needs further investigation     

Basal cell carcinoma stem cells exhibit osteogenic and chondrogenic differentiation potential

Specific cell subpopulations identified as cancer stem cells (CSCs) can be found in basal cell carcinoma (BCC). Generally, CSCs have a marked trans-differentiation potential that could potentially be used in differentiation therapies. However, there are no studies regarding BCC CSCs multipotency. The aim of the study was to analyze the characteristic of CSCs of BCC with emphasis on their differentiation potential upon specific induction. Specific staining and cell morphology were used for differentiation confirmation, along with the expression analysis of osteogenic (ALP, BSP, Runx2, OCN, BMP2), chondrogenic (COL1 and COL2A1), adipogenic (PPAR-γ) and neurogenic (Nestin and MAP2) markers. BCC CSCs differentiated into osteogenic and chondrogenic lineages, as judged by staining and high expression of specific markers (from 2-to 92-fold higher upon induction). Concomitantly with differentiation, the levels of cancer stem cell markers decreased in the cultures. Adipo-differentiation and neuro-differentiation were unsuccessful. In conclusion, BCC CSCs exhibit the capacity to trans-differentiate, a characteristic that may potentially be useful in the development of new strategies for the treatment of aggressive BCCs.     

TO LOAD BALANCING IN COMPUTER NETWORKS

Abstract

This paper presents an AI-based approach to Load balancing in computer networks; the load redistribution in the network is controlled by a rule-based expert system. Our experiments have shown that changes in the workload structure impose the necessity of automated modifications of the rules in the knowledge base. These modifications are made by means of a machine learning subsystem that is incremental and capable of forgetting pieces of knowledge that have become obsolete. This paper can be understood as a brief report on a real-world application of Michalski's idea of flexible concepts.     

Buffers and backup machines in automatic transfer lines

Buffer storage of the work-in-process inventory, decouples successive stages of automatic transfer production lines, assuring partial operability under machine failure. Certain special features such as secondary (standby) machines, special maintenance and diagnostic systems can lengtben the uptime or shorten the downtime of individual stages. When special features for K stages and spaces for M buffers are available, it is of great interest to system designers to know which stages should have the special features and where the buffer spaces should be inserted to maximize the line output rate. This paper addresses itself to such design problems. A bivariate dynamic programming procedure is developed which provides a layout for the buffers and an allocation of special features, maximizing the line output rate as defined by Buzacott (1967, 1968). The Buzacott formula is based on a heuristic argument which provides, in general, an upper bound on the true system output rate.