CHEF

نویسندگان

چکیده

High-quality labels are expensive to obtain for many machine learning tasks, such as medical image classification tasks. Therefore, probabilistic (weak) produced by weak supervision tools used seed a process in which influential samples with identified and cleaned several human annotators improve the model performance. To lower overall cost computational overhead of this process, we propose solution called CHEF (CHEap Fast label cleaning), consists following three components. First, reduce annotators, use Infl, prioritizes most training cleaning provides save one annotator. Second, accelerate sample selector phase constructor phase, Increm-Infl incrementally produce samples, DeltaGrad-L update model. Third, redesign typical pipeline so that iteratively clean smaller batch rather than big samples. This yields better over all performance enables possible early termination when expected has been achieved. Extensive experiments show our approach gives good prediction while achieving significant speed-ups.

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Chef de Clinique- Assistant des Hôpitaux

The potential role of nanotechnologies in the future of psychiatry. A systematic review. Abstract. Nanomedicine is defined as the area using nanotechnology's concepts for the benefit of human beings' health and well being. In this article, we aimed to provide an overview of areas where nanotechnology is applied and how they could be extended to care for psychiatric illnesses. The main applicati...

متن کامل

CHEF: A Model of Case-Based Planning

Case-based planning is based on the idea that a machine planner should make use of its own past experience in developing new plans, relying on its memories instead of a base of rules. Memories of past successes are accessed and modified to create new plans. Memories of past failures are used to warn the planner of impending problems and memories of past repairs are called upon to tell the plann...

متن کامل

CHEF: Convex Hull of Elliptic Features for 3D Blob Detection

We present an efficient protocol for robust detection of 3D blobs from volumetric datasets. The approach has three steps. The first step of the process detects elliptic features by classifying the Hessian of the scale space representation of the volume data. These features are then grouped into 3D connected components, which are subsequently partitioned by computing a convex hull of each connec...

متن کامل

Cloud Service Orchestration with TOSCA, Chef and Openstack

As the sector of SaaS is being evolved, the need of portability and effective orchestration of Cloud-enabled applications over virtualised infrastructures is more apparent. In this paper we present a proof-of-concept project which deals with the deployment and orchestration of Cloud applications through the 1st published version of the TOSCA specification. In our work we investigate the state o...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Proceedings of the VLDB Endowment

سال: 2021

ISSN: ['2150-8097']

DOI: https://doi.org/10.14778/3476249.3476290