STIMULATION OF COGNITIVE FUNCTIONS OF PSYCHE BY MEANS OF ONLINE GAMES

The article presents the results of testing of the developed online-training of cognitive functions of the brain, which was implemented on a separately created game platform using a set of specially developed games. The proposed results reveal the theoretical justification of the cognitive functions, which are selected for training: memory, thinking, attention, imagination, perception. The characteristics of each cognitive function to be stimulated in the conditions of online games are highlighted. We described the algorithm for games evaluating according to certain cognitive functions and their characteristics. An analysis of the effectiveness of evaluation is given. The results of completing by players of game sessions balanced on the chosen cognitive functions are described. A comparison of the effectiveness of stimulation of different cognitive functions depending on the motivation and choice of players is presented. The aim of the study was to evaluate the effectiveness of online games in activation and stimulation of the cognitive functions of the human psyche. The research methods were based on our own developed algorithm for training success estimation, analysis of the results of online games, methods of mathematical statistics. The results of the study are presented on the basis of approbation of the developed training, conducted during two years. The analysis of the implementation of the training revealed the dependence of the effectiveness of stimulation of cognitive functions on the motivation and choice of the player. It is shown that when training on a balanced set of games, where all cognitive functions are present in equal shares, the effectiveness of training is not uniform and not balanced. It is investigated that the effectiveness of training depends on the quality of the game's effect on cognitive function. In order to take into account this quality of influence, all games were divided into two classes: actively shaping and passively stimulating. The distribution of games by classes was experimentally confirmed. A generalized analysis of the game sessions of all players for six months revealed cognitive functions which for the most players were better developed and more stimulated.

Optimal memory-aware backpropagation of deep join networks

Deep learning training memory needs can prevent the user from considering large models and large batch sizes. In this work, we propose to use techniques from memory-aware scheduling and automatic differentiation (AD) to execute a backpropagation graph with a bounded memory requirement at the cost of extra recomputations. The case of a single homogeneous chain, i.e. the case of a network whose stages are all identical and form a chain, is well understood and optimal solutions have been proposed in the AD literature. The networks encountered in practice in the context of deep learning are much more diverse, both in terms of shape and heterogeneity. In this work, we define the class of backpropagation graphs, and extend those on which one can compute in polynomial time a solution that minimizes the total number of recomputations. In particular, we consider join graphs which correspond to models such as siamese or cross-modal networks. This article is part of a discussion meeting issue ‘Numerical algorithms for high-performance computational science’.

Training Transfer: Does Training Design Preserve Training Memory?

Billions of dollars are lost by low application of ineffective training. Fast declination of training memory may contribute this loss. The present study uses theoretical examinations via a conceptual model to examine the relationship between training memory and transfer behaviour. Training design, training retention (training memory), and training transfer are the study variables. The study population, is the federal ministries in the United Arab Emirates (UAE), was assessed via random sampling. Data were collected by a cross-sectional approach via questionnaires. Back-translation (English to Arabic), a pre-test, and a pilot test were applied to ensure that any modifications of the questionnaire items were precise and effective. The study was analysed via PLS-SEM. Based on the results, all of the study’s hypotheses were accepted, and significant relationships were revealed between the study variables. Training design is highly correlated with training retention, i.e., a premium training design will lead to a high preservation of the knowledge and skills gained from the training programme. Due to the low correlation between training retention and training transfer, the training retention was considered a secondary contributor of applying training to the work environment. If mangers and practitioners tend to achieve successful training transfer, their efforts should concentrate on adopting modern training design techniques, which could sufficiently maintain the training memory and increase training transfer.

Post-extinction Delay Necessary to Induce Retrograde Amnesia for a Moderate Extinction Training Memory

Two experiments using rats were conducted to evaluate the post-extinction delay necessary to obtain retrograde amnesia for a moderate extinction training memory. In Experiment 1, six minutes of extinction (i.e., cue-exposure) was sufficient to reduce fear of the black compartment of a white-black shuttle box, however the amnestic treatment cycloheximide (CHX) failed to produce retrograde amnesia for extinction (i.e., show fear). In Experiment 2, CHX was administered at various post-extinction delays (0-min, 60-min, 75-min, 120-min) to assess whether the active extinction memory could be susceptible to amnesia if the original fear memory had time to reconsolidate. The results indicated that administrating CHX 75 minutes after extinction produced retrograde amnesia for extinction, but not for shorter post-extinction delays, thus demonstrating a temporal gradient. These findings suggest that the extinction memory was active and susceptible to disruption 75 minutes after the extinction session, but the original fear memory may have been protected from the amnestic effects with sufficient time to reconsolidate.