Vertaling van "machine learning quality management" in Japans
We konden deze vermelding niet vinden. Er worden benaderende resultaten weergegeven. Controleer je spelling of stel voor deze term aan het woordenboek toe te voegen.
機械学習品質マネジメント
From the viewpoint of machine learning quality management, we focus especially on changing characteristics such as the status of hazard and risks.
機械学習品質マネジメントの観点 からは、特に危害の状況やリスクなどが変化するような特徴に着目する。
(Example 2) When image recognition AI that aims to recognize "cats" from small animals around town is to be built and individual characteristics of cats such as breed and size are excluded from recognition, this characteristic means that sufficiently diverse images of "cats" and "other small 18 link to page 26 link to page 24 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology animals such as dogs" are made available as data and that training does not only target specific breeds (e.g. calico cats) in environments where the product is expected to be used.
(例 2) 20 link to page 29 link to page 27 機械学習品質マネジメントガイドライン 第 1 版 国立研究開発法人産業技術総合研究所 街中の小動物から猫を認識する画像認識 AI を構築する際に、猫の品種や体長 などを個別の特徴として認識しないこととした際に、その製品が利用が想定される 環境において十分に多様な猫や犬など他の小動物の画像がデータとして用意 されること、例えば三毛猫など特定の品種だけが学習対象になっていないことが、こ の特性に当たる。
In an extremely simple system, it is enough to mention that corresponding data to all 17 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology "combinations of situations" identified in the problem domain analysis described in the previous section included in training datasets and test datasets.
Robustness This characteristic allows systems to maintain their performance level under any circumstance. 34 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology Terms related to development process 1. System lifecycle process A process management model looking down the flow from the planning of systems to the end of operation and disposal.
堅牢性 robustness システムが性能のレベルをどのような状況下でも維持できる性質。 38 機械学習品質マネジメントガイドライン 第 1 版 国立研究開発法人産業技術総合研究所 開発プロセスに関する用語 1. システムライフサイクルプロセス system lifecycle process システムの企画立案から運用終了廃棄までの一連の流れを俯瞰した工程管理モデル。
27 link to page 35 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology 2 Overview 2.1 Scope of the Guideline Products and systems subject to the Guideline Products and services subject to quality management in this Guideline are those that use machine learning technology for the building of some of their software components among all information-processing systems such as industrial products, consumer equipment and information services (hereinafter referred to as "machine-learning based system" in this Guideline).
30 link to page 39 機械学習品質マネジメントガイドライン 第 1 版 国立研究開発法人産業技術総合研究所 2 基本的事項 2.1 ガイドラインのスコープ 対象とする製品・システム 本ガイドラインが品質マネジメントの対象とする製品・サービスは、産業製品・消費者機 器・情報サービスなど情報処理を行うシステム全般のうちで、内包される一部のソフトウェ ア要素の構築に機械学習技術を用いたもの(以下、本ガイドラインにおいて機械学習利用 システムという。
46 link to page 51 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology Data cleaning Noise removal, data forming, complement of missing values and removal of outliers are carried out so that original characteristics of selected datasets can be trained.
53 link to page 59 機械学習品質マネジメントガイドライン 第 1 版 国立研究開発法人産業技術総合研究所 データクリーニング 選別したデータセットに対して、本来の特徴を訓練できるように、ノイズ除去、データ整 形、欠損値補完、外れ値除去などを行う。
50 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology 5 How to apply this guideline 5.1 Basic application process In this section, we overview the process of applying the Guideline.
Access Sitemap Japanese Home News About CPSEC Achievements Team Events Recruit Contact HOME Achievements Machine Learning Quality Management Guideline Committee for machine learning quality management About "Committee for machine learning quality management" "Committee for machine learning quality management" is a joint activity among researchers and engineers in AIST, academia and industry partners, to produce guidelines, methodology, technology and software for quality management of machine learning-based AI.
Access Sitemap Japanese Home News About CPSEC Achievements Team Events Recruit Contact HOME Achievements Machine Learning Quality Management Guideline Machine Learning Quality Management Guideline CPSEC and AIRC (Artificial Intelligence Research Center) of AIST jointly publish "Machine Learning Quality Management Guideline". This document establishes a basis for quality goals for machine learning-based products/services, and provides procedural guidance for realizing quality through development process management and system evaluations.
1.3 Chal enges behind machine learning quality management The implementation of artificial intelligence by machine learning is, if simply said, a sort of software.
When the test environment is different from actual operation environments, it is necessary to evaluate the differences between them. In the development of machine learning systems, however, there are often significant differences between an environment used for training (e.g. computing environment with cloud and GPU) 74 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology and an actual operational environment (e.g. built-in computer).
Second, at the same time, it is necessary to cover as much as possible all situations which the machine learning based system to be implemented 64 link to page 22 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology may encounter during operation from the viewpoint of both the quality of training and the quality of deliverables.
8 link to page 14 link to page 15 link to page 15 Machine Learning Quality Management Guideline National Institute of 1st English edition Advanced Industrial Science and Technology (Supplement) When a product or service is developed using strict processes to evaluate safety and reliability of the whole industrial systems such as IEC 61508 4 or IEC 62278 5, a requirement for external qualities is of course identified in relation to machine learning components in the design process of whole conventional systems.
Er worden alleen voorbeelden gegeven om u te helpen het woord of de woordcombinatie waarop u hebt gezocht, te vertalen. Deze worden niet door ons geselecteerd of gevalideerd en kunnen ongepaste taal bevatten. Wij vragen u melding te maken van voorbeelden die dienen te worden aangepast of verwijderd. Vertalingen met grof of informeel taalgebruik worden meestal rood of oranje gemarkeerd.