China 
Africa 
Explore chapters and articles related to this topic
The Value Stream–Choosing Programming Languages and Tools
Published in Peter Middleton, James Sutton, Lean Software Strategies, 2020
On the other hand, Java adds features that are hard on integrity and that aren’t found in C++, such as automatic garbage collection (memory cleanup), dynamic class loading mechanisms, object serialization (which permits a kind of limited backdoor access to private data), the Java Memory Model (with semantic difficulties), and many factors leading to nondeterministic (i.e., unpredictable) execution timing. “The construction of high-integrity real-time systems requires that both the functional and the temporal behavior of applications is analyzable. This requirement . . . influences the semantics that the language of the profile has to provide.”17 These factors have seriously degraded Java’s integrity: “Despite all [its] valuable features, Java has been criticized for its unpredictable performance as well as some security concerns . . . [several of its] mechanism are often considered problematic, especially under time- or performance-critical situations. Moreover, a number of security bugs in the Java virtual machine have been discovered . . . . These fears make Java and its associated technology simply unsuitable for high-integrity systems.”18
Application of an Integrated System for Process and Fixture Planning
Published in Awais Ahmad Khan, Emad Abouel Nasr, Abdulrahman Al-Ahmari, Syed Hammad Mian, Integrated Process & Fixture Planning, 2018
Awais Ahmad Khan, Emad Abouel Nasr, Abdulrahman Al-Ahmari, Syed Hammad Mian
The methodology presented is successfully tested and validated with multiple fixturing setup plans. All the relevant geometrical data is successfully extracted from the STEP AP203 file. The EXPRESS classes are used to describe entities in STEP AP203 and map them into the C++ class’s information memory model. This information identifies every B-rep solid using its faces, loops, edges, and vertices along with the surface type and the normal vector direction. The faces, edges, and vertices have been given a unique id number. Information regarding the face conditions (plane, cylindrical, and conical), edge conditions (plane line, tangent line, and circle edge), and face and edge directions are collected. Moreover, the orientation, direction of a face vector, and external and internal loop determination are also established. The geometric database for the part is created based on an object-oriented technique.
Introduction
Published in Joseph Y.-T. Leung, Handbook of SCHEDULING, 2004
We describe here algorithms and lower bound results for multiprocessor scheduling of overloaded real-time systems. We consider two memory models: a shared memory model where thread migration is cheap, and a distributed memory model where thread migration is impractical. In both cases, we assume a centralized scheduler. In the first model, tasks can migrate cheaply (and quickly) from one processor to another. Hence, if a task starts to execute on one processor it can later continue on any other processor (and migration takes no time). In the second model (the fixed model), once a task starts to execute on one processor it cannot execute on any other processor. For both models, we assume that preemption within a processor takes no time. Main results are given below: Inherent Bound on the Best Possible Competitive Multiplier
An Agglomerative Hierarchical Clustering Framework for Improving the Ensemble Clustering Process
Published in Cybernetics and Systems, 2022
Mohammad Jafarzadegan, Faramarz Safi-Esfahani, Zahra Beheshti
The LIMBO clustering algorithm is a hierarchical clustering algorithm that uses the concept of Bioinformatics Clustering (BI) to calculate the distance between batch traits (Andritsos et al. 2004). The advantage of this algorithm is the production of clusters of different sizes in one implementation of the algorithm. LIMBO manages a large data set by generating a finite memory model. This algorithm also manages distorted and noisy data well. The COOLCAT clustering algorithm uses the concept of entropy to group records (Barbará, Li, and Couto 2002). This algorithm is an incremental algorithm with the aim of minimizing the expected entropy for the clusters. Having a set of clusters, the COOLCAT algorithm clusters the next point in the data point set by minimizing the expected total entropy. This algorithm performs clustering without any preprocessing on the dataset. Therefore, COOLCAT is suitable for streaming data but does not have the ability to manage outbound data. Miller et al. (2021) proposed a deterministic algorithm called EREW for AHC, which is based on a complete graph and spanning tree algorithms with a minimum Euclidean distance. This algorithm can cluster objects at time
A hardware intelligent processing accelerator for domestic service robots
Published in Advanced Robotics, 2020
Yutaro Ishida, Takashi Morie, Hakaru Tamukoh
The computational resources of the embedded CPU are very limited; thus, on the embedded CPU, data exchange requires a communication model with a smaller computational load than that of the ROS interface, as illustrated in Figure 2. Moreover, the ROS interface is not well known among circuit engineers. Therefore, we propose a shared memory communication model instead of using the ROS interface. This model exchanges data by utilizing only internal memory; therefore, its computation load is less than that of the ROS interface, which exchanges data through a network. The model is implemented by an inter-process communication on UNIX, using ‘shmget ,' ‘shmat,’ ‘ shmdt’ and ‘shmctl’ functions. As shown in Figure 4(a), an SHM object implemented in the model provides an interface to the FPGA controller. This interface is compatible with the ‘memcpy’ and ‘memset’ functions, in the C language, circuit engineers are familiar with. Additionally, as shown in Figure 4(b), the object provides an interface to the ROS space/other processes that is compatible with ‘publisher’ and ‘subscriber ,' which robotic engineers are familiar with, in ROS. Through the interface, the model can exchange the data of both circuit and robotic engineers through the shared memory model.
A comparative evaluation of machine learning algorithms and an improved optimal model for landslide susceptibility: a case study
Published in Geomatics, Natural Hazards and Risk, 2021
Yue Liu, Peihua Xu, Chen Cao, Bo Shan, Kuanxing Zhu, Qiuyang Ma, Zongshuo Zhang, Han Yin
The ANN model was constructed in Matlab (2020a, MathWorks) with landslides and non-landslides pixel using memory-based learning. The input training and test data were employed to establish a memory model and the results were stored in a large memory source. Following the instruction of a new test vector, the learning process places the new vector into a stored class for classifications (Choi et al. 2012; Dou et al. 2015). Here 50% training data and 50% test data had been experimented but did not perform perfectly. Therefore, 30% and 70% of the test and training data were adopted, respectively. Figure 9 displays the contribution of each factor, with lithology, aspect, and elevation observed as the dominant factors impacting landslides.