Pointer-Based Divergence Analysis for OpenCL 2.0 Programs

A modern GPU is designed with many large thread groups to achieve a high throughput and performance. Within these groups, the threads are grouped into fixed-size SIMD batches in which the same instruction is applied to vectors of data in a lockstep. This GPU architecture is suitable for applications with a high degree of data parallelism, but its performance degrades seriously when divergence occurs. Many optimizations for divergence have been proposed, and they vary with the divergence information about variables and branches. A previous analysis scheme viewed pointers and return values from functions as divergence directly, and only focused on OpenCL 1.x. In this article, we present a novel scheme that reports the divergence information for pointer-intensive OpenCL programs. The approach is based on extended static single assignment (SSA) and adds some special functions and annotations from memory SSA and gated SSA. The proposed scheme first constructs extended SSA, which is then used to build a divergence relation graph that includes all of the possible points-to relationships of the pointers and initialized divergence states. The divergence state of the pointers can be determined by propagating the divergence state of the divergence relation graph. The scheme is further extended for interprocedural cases by considering function-related statements. The proposed scheme was implemented in an LLVM compiler and can be applied to OpenCL programs. We analyzed 10 programs with 24 kernels, with a total analyzed program size of 1,306 instructions in an LLVM intermediate representation, with 885 variables, 108 branches, and 313 pointer-related statements. The total number of divergent pointers detected was 146 for the proposed scheme, 200 for the scheme in which the pointer was always divergent, and 155 for the current LLVM default scheme; the total numbers of divergent variables detected were 458, 519, and 482, respectively, with 31, 34, and 32 divergent branches. These experimental results indicate that the proposed scheme is more precise than both a scheme in which a pointer is always divergent and the current LLVM default scheme.

Filippova Automated planning of graduality assembly

Modern industrial production requires the improvement of assembly processes, and thus increase the level of automated intelligent sequence planning. Therefore, researches in the field of automation of the sequence of assembly of products in industries are relevant at this time. In today's world there is a need to develop complex, accurate products. Problems are created in industries due to the reduction of the life cycle of products. There is a need to study the problem of assembly planning to achieve the goal of practical implementation and standardization of assembly plans. Creating graphs of the addition process is one of the problems. The assembly planning system can reduce human intervention in the process and reduce computational effort. The finished assembly contains many components that can be assembled using many sequences. A review of the methods from the literature showed that although these methods increase the automation level, they still cannot be applied to actual production because they do not take into account the experience and knowledge that can play a major role in planning and are of great value. Assembly planning, relationship charts, priority charts. Improving the assembly planning system to create a communication schedule and an assembly priority schedule was proposed. The advanced system will be used to generate possible assembly sequences with subassembly identification. A system has been developed to create alternative possible assembly sequences that can be used by component part / product designers in the early stages. A system capable of generating assembly sequences for simultaneous assembly of multiple parts has been proposed. Conclusions and work results can be applied used and improved for more productive product development by designers in the early stages and faster assembly of products in enterprises. The paper did not consider practical limitations (gravity) and irreversible assembly operations, such as permanent fastening, welding etc. Кey words: assembly, blocking graph, relation graph, sequence

Deep Reinforcement Learning for Stock Recommendation

Recommending stocks is very important for investment companies and investors. However, without enough analysts, no stock selection strategy can capture the dynamics of all S&P 500 stocks. Nevertheless, most existing recommending strategies are based on predictive models to buy and hold stocks with high return potential. But these strategies fail to recommend stocks from different industrial sectors to reduce risks. In this article, we propose a novel solution that recommends a stock portfolio with reinforcement learning from the S&P 500 index. Our basic idea is to construct a stock relation graph (RG) which provide rich relations among stocks and industrial sectors, to generate diversified recommendation result. To this end, we design a new method to explore high-quality stocks from the constructed relation graph with reinforcement learning. Specifically, the reinforcement learning agent jumps from each industrial sector to select stock based on the feedback signals from the market. Finally, we apply portfolio allocation methods (i.e., mean-variance and minimum-variance) to test the validity of the recommendation. The empirical results show that the performance of portfolio allocation based on the selected stocks is better than the long-term strategy on the S&P 500 Index in terms of cumulative returns.

A Multi-graph Convolutional Network Framework for Tourist Flow Prediction

With the advancement of Cyber Physic Systems and Social Internet of Things, the tourism industry is facing challenges and opportunities. We can now able to collect, store, and analyze large amounts of travel data. With the help of data science and artificial intelligence, smart tourism enables tourists with great autonomy and convenience for an intelligent trip. It is of great significance to make full use of these massive data to provide better services for smart tourism. However, due to the skewed and imbalanced visiting for point of interest located at different places, it is of great significance to predict the tourist flow of each place, which can help the service providers for designing a better schedule visiting strategy in advance. Against this background, this article proposes a multi-graph convolutional network framework, named AMOUNT, for tourist flow prediction. To capture the diverse relationships among POIs, AMOUNT first constructs three subgraphs, including the geographical graph, interaction graph, and the co-relation graph. Then, a multi-graph convolution network is utilized to predict the future tourist flow. Experimental results on two real-world datasets indicate that the proposed AMOUNT model outperforms all other baseline tourist flow prediction approaches.

Keywords-Driven and Weight-aware Paper Recommendation via Paper Correlation Pattern Mining

Nowadays, recommender systems have become one of the main tools and methods for users to search for their interested papers from massive candidates. Considering the above drawbacks, in this paper, we propose a link prediction approach that combines time, keywords and authors information for constructing a new relation graph. Finally, a case study is employed to explain our approach step by step and demonstrate the feasibility of our proposal.

Graph-Based Stock Recommendation by Time-Aware Relational Attention Network

The stock market investors aim at maximizing their investment returns. Stock recommendation task is to recommend stocks with higher return ratios for the investors. Most stock prediction methods study the historical sequence patterns to predict stock trend or price in the near future. In fact, the future price of a stock is correlated not only with its historical price, but also with other stocks. In this article, we take into account the relationships between stocks (corporations) by stock relation graph. Furthermore, we propose a Time-aware Relational Attention Network (TRAN) for graph-based stock recommendation according to return ratio ranking. In TRAN, the time-aware relational attention mechanism is designed to capture time-varying correlation strengths between stocks by the interaction of historical sequences and stock description documents. With the dynamic strengths, the nodes of the stock relation graph aggregate the features of neighbor stock nodes by graph convolution operation. For a given group of stocks, the proposed TRAN model can output the ranking results of stocks according to their return ratios. The experimental results on several real-world datasets demonstrate the effectiveness of our TRAN for stock recommendation.