使用期限租赁和永久
许可形式单机和网络版
原产地美国
介质下载
适用平台windows,mac,linux
科学软件网提供大量正版科学软件,满足各学科的科研要求。科学软件网专注软件销售服务已达19年,全国大部分高校和企事业单位都是我们的客户。同时,我们还提供本地化服务,助力中国的科研事业。
Acquire live data from individual instruments, cards, sensors, or internet of things approaches
Read or write stored data including signals, images, video, and ‘omics data from files, databases, spreadsheets, or via web access
Manage datasets too large to fit in memory
Model and simulate biological systems with an intuitive graphical interface or symbolic math
Refine your models with optimization, curve fitting, and parameter estimation
Design experiments and characterize results with frequentist or Bayesian statistics or machine learning
Report results
Scale up computing to multicore machines and GPUs, clusters and HPC centers, and the cloud
Collaborate and support teams with a range of deployment options including apps, desktop or web executables, and mobile devices
Deploy royalty-free integrations with third-party software and programming languages
Re-use code assets developed and shared by the global scientific community
Readily automate tasks into pipelines or processes using code that can be shared and evolved
MathWorks also provides training and consulting services to help you or your team become more proficient and complete projects faster.
Because MATLAB and Simulink® toolboxes have been trusted by the global science community for over 30 years, researchers and educators have created a large and diverse collection of domain-specific tools written in MATLAB. Many of these tools are freely available at MathWorks File Exchange, GitHub, or the MathWorks Connections Program.
为什么使用 MATLAB 实现深度学习?
互操作性
不再有框架基于 MATLAB 还是 Python 的选择题。使用 ONNX 导入和导出功能,MATLAB 支持与开源深度学习框架的互操作性。使用 MATLAB 工具的大意义在于——访问 Python 中没有的功能与预置函数及应用程序。
预处理应用程序
快速开始网络训练。使用特定领域应用程序快速预处理音频、视频和图像数据集。使用 Deep Network Designer 应用程序创建复杂的网络架构,或修改预训练网络以进行迁移学习,在训练之前可视化、检查并修复问题。
多平台部署
可随处部署深度学习模型,包括 CUDA、C 代码、企业系统或云。若在意性能,您可以利用 Intel® (MKL-DNN)、NVIDIA(TensorRT、cuDNN)和 ARM® (ARM Compute Library) 优化库生成代码,创建具有高性能推理速度的可部署模型。
从 MATLAB 生成代码
从 MATLAB 生成 C、C++、CUDA 和 HDL 代码,只需三个迭代步骤。从任意位置运行和部署代码,台式机、移动设备或嵌入式系统均可。MATLAB Coder 可以随着设计的推进自动生成新代码,无需手工编码。这样可以加速并实现更多的设计迭代。
您可以将生成的代码作为源代码、静态库或动态库,集成到桌面或云端中脱离 MATLAB 环境运行的应用程序。您还可以将生成的代码打包成 MEX 函数并直接在 MATLAB 中使用。
从 Simulink 生成代码
使用 Embedded Coder®,将您的模型转换为高质量源代码和可执行文件,从而实现原型设计和生产。既可使用默认 C 和 C++ 设置,也可使用 SIMD 指令、数据存储类及内存区段代码放置进行优化以实现更高性能。生成 ANSI/ISO C/C++、MISRA-C 或 AUTOSAR 源代码,插入运行时系统。或者,生成完整的交钥匙可执行文件,自动下载并在您的自定义硬件设备上运行。
使用双向可追踪链接,检查模型的生成代码,或使用代码、数据、函数接口和代码指标,生成报告。
Accelerate Analysis with High Performance Computing
MathWorks continues to accelerate large computations for big data problems by supporting the latest multithreaded hardware architectures. Using Parallel Computing Toolbox™, MATLAB Parallel Server™, and or MATLAB Production Server™, computer scientists and data analysts are getting answers faster than ever before by utilizing the high-performance computing power of multicore processors, GPUs, and compute farms.
Deploy and Integrate Data Analytics Models into Enterprise Applications
While MATLAB offers an environment for developing advanced data analysis and machine learning algorithms, these models and systems often must be deployed into the real world. MathWorks provides many different options for the deployment of MATLAB algorithms including generation of portable C/C++ code, compiled executables, web-based applications, or even integration to mobile apps.
19年来,公司始终秉承、专注、专心的发展理念,厚积薄发,积累了大量的人才、技术以及行业经验,在行业内得到了大量用户的认可和高度价。
http://www.kxrjsoft.com.cn