Evaluation of software clustering algorithms is typically done by comparing the clustering results to an authoritative decomposition prepared manually by a system expert. A well-known drawback of this approach is the ...
详细信息
Evaluation of software clustering algorithms is typically done by comparing the clustering results to an authoritative decomposition prepared manually by a system expert. A well-known drawback of this approach is the fact that there are many, equally valid ways to decompose a software system, since different clustering objectives create different decompositions. Evaluating all clustering algorithms against a single authoritative decomposition can lead to biased results. In this paper, we introduce and quantify the notion of clustering algorithm comparability. It is based on the concept that algorithms with different objectives should not be directly compared. Not surprisingly, we find that several of the published algorithms in the literature are not comparable to each other.
Selecting an appropriate software clustering algorithm that can help the process of understanding a large software system is a challenging issue. The effectiveness of a particular algorithm may be influenced by a numb...
详细信息
Selecting an appropriate software clustering algorithm that can help the process of understanding a large software system is a challenging issue. The effectiveness of a particular algorithm may be influenced by a number of different factors, such as the types of decompositions produced, or the way clusters are named. In this paper, we introduce an effectiveness measure for software clustering algorithms based on Mojo distance, and describe an algorithm that calculates its value. We also present experiments that demonstrate its improved performance over previous measures, and show how it can be used to assess the effectiveness of software clustering algorithms.
In this paper, we report and share our experience on introducing the teaching of software Testing in an algorithms and Data Structures course. The goal was twofold: (1) help in the teaching of the complex subject of a...
详细信息
ISBN:
(数字)9781728110752
ISBN:
(纸本)9781728110769
In this paper, we report and share our experience on introducing the teaching of software Testing in an algorithms and Data Structures course. The goal was twofold: (1) help in the teaching of the complex subject of algorithms and data structures and (2) start to make the student familiar with the topic of software testing. We report on the lessons learned in such endeavor and on the results of a student survey, after the course was completed. How the teaching of software testing was introduced was positively evaluated by the students. All our teaching material is freely available online on GitHub.
作者：
L.W. HendersonKnobbe
Martens Olson and Bear Limited Liability Partnership Newport Beach CA USA
In the US, software has been patentable subject matter for many years. Nevertheless, it was not uncommon to hear engineers, and even some patent attorneys, assert that software was not patentable. The patentability of...
详细信息
In the US, software has been patentable subject matter for many years. Nevertheless, it was not uncommon to hear engineers, and even some patent attorneys, assert that software was not patentable. The patentability of mathematical algorithms in the US is not so clear cut, but changes in US patent law have expanded the scope of patentability for mathematical algorithms. Many types of mathematical algorithms that were previously not patentable subject matter are now patentable subject matter. The full impact of these changes is yet to be determined, but the trend is clearly in favor of the patentability of mathematical algorithms. In Europe, the expanding scope of patentability for software and mathematical algorithms is lagging the US to some extent. The patentability of software in Europe was somewhat uncertain. However, a decision by the Technical Board of Appeal (TBA) appears to have removed the uncertainty and clearly established that computer software is generally patentable subject matter in the European Patent Office (EPO). The patentability of a mathematical algorithm is still uncertain. Inventors wanting to obtain patent protection for their mathematical algorithms or computer software should not assume that such protection is unavailable or even that they know where the boundary lies. It is a mistake, and it can be a costly mistake, to presuppose that something is unpatentable simply because it is a mathematical algorithm.
The quick expansion of software development necessitates effective tools for automating monotonous coding jobs, yet existing solutions frequently fall short in producing accurate codes from natural language explanatio...
详细信息
ISBN:
(数字)9798331512248
ISBN:
(纸本)9798331512255
The quick expansion of software development necessitates effective tools for automating monotonous coding jobs, yet existing solutions frequently fall short in producing accurate codes from natural language explanations, resulting in a disconnect between machine execution and human intent. This project uses cutting-edge AI methods to generate Python code automatically in order to overcome this difficulty. A dataset of Python scripts is used to train the model of a transformer with a vanilla encoder-decoder architecture. To improve the model's comprehension of natural language inputs, the data is preprocessed using a variety of embedding approaches. The model can accurately translate descriptions of English text into Python code. software development is revolutionized by machine learning and predictive algorithms when the model that has been trained is then made available through a streamlit UI, which allows for real-time, user friendly engagements for code production based on English language inputs. The goal of this effort is to create an AI-powered system that uses a transformer-based vanilla encoder-decoder paradigm to accurately translate descriptions from natural English into Python code. Through the use of predictive algorithms and sophisticated embedding techniques, the system seeks to increase software development automation, increasing productivity and decreasing the amount of manual coding required.
LibRA is a scientific software library with implementations of both existing and new algorithms for interfer-ometric calibration and imaging in the Radio Astronomy (RA) community. Derived from the CASA code base, LibR...
详细信息
ISBN:
(数字)9781946815200
ISBN:
(纸本)9798331513979
LibRA is a scientific software library with implementations of both existing and new algorithms for interfer-ometric calibration and imaging in the Radio Astronomy (RA) community. Derived from the CASA code base, LibRA has been developed into a standalone, open-source library with its own independent reposi-tory. The primary goal of the LibRA project is to provide direct access to RA algorithms, targeting research groups, university departments, and individual researchers, including faculty, students, and postdoctoral re-searchers. This library offers production-quality implementations of core algorithms that are user-friendly, easily deployable, and modifiable to suit specific needs. The fundamental scientific principles behind such telescopes share commonalities with other domains that rely on indirect imaging, such as Magnetic Res-onance Imaging (MRI) and ultrasound imaging. To make RA algorithms accessible, a strong requirement driving the software design is to keep the interface simple, the software stack shallow, and the software de-pendency graph small. To achieve these primary goals, LibRA provides interfaces through C++, Python, and command-line options. In this talk, we will outline the software architecture, current capabilities, and in-terfaces of LibRA. We will also discuss how these can be leveraged in other software development projects and by end-users for scientific applications. Additionally, we will illustrate the use of LibRA by parallel computing in large-scale distributed networks for calibration and imaging of ngVLA/SKA-sized data sets. The talk will also cover ongoing collaborations with computing, RA, and medical imaging groups, as well as future development plans.
The article presents a description of the developed software platform that provides automation of data storage server state monitoring using a hybrid algorithm. The developed algorithm combines an ontological approach...
详细信息
ISBN:
(数字)9798331511241
ISBN:
(纸本)9798331511258
The article presents a description of the developed software platform that provides automation of data storage server state monitoring using a hybrid algorithm. The developed algorithm combines an ontological approach and machine learning elements. A comparative assessment of various approaches was carried out: cluster analysis, autoencoders, LSTM, fuzzy ontologies and hybrid models - when analyzing methods for searching for anomalies in time series. The hybrid evaluation model showed the best results for all criteria. It was chosen to develop a prototype of the software platform. Experiments were conducted on hardware storage system data. The software platform proved its effectiveness in solving the problem.
In recent years, maritime surveillance of ships using High Frequency Surface Wave Radars (HFSWRs) represents very important challenge, where most attention is focused on algorithms for primary signal processing and th...
详细信息
ISBN:
(数字)9798331515799
ISBN:
(纸本)9798331515805
In recent years, maritime surveillance of ships using High Frequency Surface Wave Radars (HFSWRs) represents very important challenge, where most attention is focused on algorithms for primary signal processing and the implementation of such algorithms to make them applicable in real-time. This paper contains an overview of the high-resolution primary signal processing in HFSWRs, as well as its implementation using C programming language. All the necessary input data and parameters of the algorithms are described, then the way of executing the program, and finally the output data. A real-time operation is achieved thanks to the implementation based on multithreading by improving the existing implementation of the high-resolution range-Doppler map estimation, which is the most computationally demanding task. As a part of software infrastructure, we have developed a set of auxiliary processing components, allowing end-users graphical presentation of processed radar data. We have demonstrated the capabilities of the ship detection software which can be execute on general purpose computers, and where processing can be performed in real-time or subsequently, for the needs of which an additional program has been developed, thus forming a ship detection software package.
The article considers methods for measuring parameters of communication links connecting OpenFlow switches in software defined networks of the Internet of Things to ensure the operation of dynamic routing algorithms. ...
详细信息
ISBN:
(数字)9798331511241
ISBN:
(纸本)9798331511258
The article considers methods for measuring parameters of communication links connecting OpenFlow switches in software defined networks of the Internet of Things to ensure the operation of dynamic routing algorithms. Theoretical information on the parameters of communication links that ensure quality of service in the network and methods for measuring them using a network controller are given. Step-by-step algorithms for implementing measurement of delay, packet loss percentage, bandwidth and load of communication links using the Ryu controller are given. Experimental studies confirming the operability of the presented algorithms are conducted. A comparison of the proposed algorithms with the operation of other utilities for measuring communication links parameters, such as Ping and Iperf3, is carried out. It is shown that the quality of the proposed algorithms for measuring delay and packet loss percentage using the controller show similar results to the quality of the Ping and Iperf3 utilities. The limitations imposed on the operation of the presented algorithm for measuring the bandwidth of communication links are shown. This study is an integral part of the project on developing intelligent algorithms for dynamic routing in software defined networks.
暂无评论