• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.1-17
• /
• 2017

The deep learning framework is software designed to help develop deep learning models. Some of its important functions include "automatic differentiation" and "utilization of GPU". The list of popular deep learning framework includes Caffe (BVLC) and Theano (University of Montreal). And recently, Microsoft's deep learning framework, Microsoft Cognitive Toolkit, was released as open-source license, following Google's Tensorflow a year earlier. The early deep learning frameworks have been developed mainly for research at universities. Beginning with the inception of Tensorflow, however, it seems that companies such as Microsoft and Facebook have started to join the competition of framework development. Given the trend, Google and other companies are expected to continue investing in the deep learning framework to bring forward the initiative in the artificial intelligence business. From this point of view, we think it is a good time to compare some of deep learning frameworks. So we compare three deep learning frameworks which can be used as a Python library. Those are Google's Tensorflow, Microsoft's CNTK, and Theano which is sort of a predecessor of the preceding two. The most common and important function of deep learning frameworks is the ability to perform automatic differentiation. Basically all the mathematical expressions of deep learning models can be represented as computational graphs, which consist of nodes and edges. Partial derivatives on each edge of a computational graph can then be obtained. With the partial derivatives, we can let software compute differentiation of any node with respect to any variable by utilizing chain rule of Calculus. First of all, the convenience of coding is in the order of CNTK, Tensorflow, and Theano. The criterion is simply based on the lengths of the codes and the learning curve and the ease of coding are not the main concern. According to the criteria, Theano was the most difficult to implement with, and CNTK and Tensorflow were somewhat easier. With Tensorflow, we need to define weight variables and biases explicitly. The reason that CNTK and Tensorflow are easier to implement with is that those frameworks provide us with more abstraction than Theano. We, however, need to mention that low-level coding is not always bad. It gives us flexibility of coding. With the low-level coding such as in Theano, we can implement and test any new deep learning models or any new search methods that we can think of. The assessment of the execution speed of each framework is that there is not meaningful difference. According to the experiment, execution speeds of Theano and Tensorflow are very similar, although the experiment was limited to a CNN model. In the case of CNTK, the experimental environment was not maintained as the same. The code written in CNTK has to be run in PC environment without GPU where codes execute as much as 50 times slower than with GPU. But we concluded that the difference of execution speed was within the range of variation caused by the different hardware setup. In this study, we compared three types of deep learning framework: Theano, Tensorflow, and CNTK. According to Wikipedia, there are 12 available deep learning frameworks. And 15 different attributes differentiate each framework. Some of the important attributes would include interface language (Python, C ++, Java, etc.) and the availability of libraries on various deep learning models such as CNN, RNN, DBN, and etc. And if a user implements a large scale deep learning model, it will also be important to support multiple GPU or multiple servers. Also, if you are learning the deep learning model, it would also be important if there are enough examples and references.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.14 no.2
• /
• pp.221-233
• /
• 2019

In Korea, small and medium sized domestic enterprises(SMEs) play an pivotal role in the national economy, accounting for 99.9% of all enterprises, 87.9% of total employment, and 48.3% of production. and SMEs was driving a real force of the development of national economy in many respects such as innovation, job creation, industrial diversity, balanced regional development. Despite their crucial role in the national development, most of SMEs suffer from a lack of R&D capabilities and equipments as well as funding capacity. Public R&D institutes can provide SMEs with valuable supplementary technological knowledge and help them build technological capacity. so, In order to effectively support SMEs, government and public R&D institutes must be a priority to know about the factors influencing the performance related to technology transfer and technological collaborations. In particular, SMEs are not only taking up a large portion of the national economy, but also their influence in politics and economy so strong that raising the competitiveness of small and medium-sized companies is a national policy goal that must be achieved in order to achieve sustained economic growth. For this reason, it is necessary to look specifically at the relationship between concepts such as the environment, strategy, and organizational culture surrounding the enterprise to enhance the competitiveness of SMEs. The paper analyzes 665 companies to find out which organizational culture affects their performance by classification and type of business of SMEs. This study demonstrated that when SMEs seek consistency in their external environment, strategies, and organizational structure to maintain their continued competitiveness. According to three-way analysis of variance (3-way ANOVA) indicates that classification of industries in SMEs has statistically significant main effects, but the type of business and organizational culture do not have significant effects. However, the company's organizational performance (operating profit) of SMES were found to differ significantly in comparison between groups according to classification standards of industries, and therefore adopted some parts. In addition, an analysis of the effect of interaction between the three independent variables of small and medium-sized enterprises has shown that there are statistically significant interaction effects among classification, types of business, and organizational cultures. The results shows that there is an organizational culture suitable for each industry classification and type of business of an entity, and is expected to be used as a basis for establishing promotion policies related to the incubation and commerciality of small and medium-sized venture companies in the future.

• (The)Study of the Eastern Classic
• /
• no.71
• /
• pp.99-128
• /
• 2018

Zhōuyì周易 is intended to explain the affairs of human beings by observing the images and works of all things in the universe, abstracting them into the $b{\bar{a}}gu{\grave{a}}$八卦, calculating the process and inducing the outcome by the method of stalk divination, in which this paper finds the origin of natural scientific thought of Zhōuyì. The way of Zhōuyì's thought on the natural science is distinguished from that of the Western's. In the West, people dismantled the objects into the parts until they reached the atom and analyzed them by the principle of causality to draw an axiomatic truth. In the meantime Zhōuyì observed and studied the dynamic functions and changes of all things for the convergence of the whole. While the way of Zhōuyì's thinking could have not contributed to the development of modern scientific development, that of the West overwhelmed Asian development passing through the period of enlightenment during 16-17 century. This paper tries to articulate the points where Zhōuyì can share its theory with the contemporary science by finding the traces of scientific thoughts in Zhōuyì. It encounters its ground from the methodology of natural science and scientific statements proposed by Zhōuyì. The essential concepts of Zhōuyì are induced from all things in nature. This can be considered as the idea of '法自然'(emulating the patterns and examples from nature). Also they observed the images and changes seen by the habits of animals, plants and human beings to sense and perceive their laws. These are regarded as the methodology of natural science in Zhōuyì. As a book of divination, the way of stalk divination is designed to calculate the future by using the system of 'numbers'. 'tàijí太極', ' yīnyáng陰陽', 'four symbols四象', '$b{\bar{a}}gu{\grave{a}}$八卦' and 'wǔxíng五行' are the essential concepts of Zhōuyì to represents the dynamic phenomena and changes of the natural order. Among them '$b{\bar{a}}gu{\grave{a}}$八卦' is a presentment to explain the structure of the world not by the individual analysis of things but by the unification of the whole through the contradictions and interchanges among them to reach the new orders. As of now, the studies of Zhōuyì in Korea have focused on the traditional perspectives, such as political and ethical philosophy. Some of recent studies, having interpreted Zhōuyì with scientific inclination have generated controversy 'Can Zhōuyì be a science?', for which scholars have hard time to reach the agreement. This paper tries to find the headwaters of the contemporary natural science by elaborating the methodology of natural science stated in Zhōuyì.