NCP-ADS NVIDIA-Certified Professional: Accelerated Data Sciencenvidia.cn NVIDIA 认证专业级加速数据科学家认证

更新时间：试题数量：购买人数：提供作者：

有效期：个月

章节介绍：共有个章节

收藏

我的练习

我的错题
(0道)

我的收藏
(0道)

我的斩题
(0道)

我的笔记
(0道)

专项练习

顺序练习 练习进度 0 / 0

随机练习 自定义设置练习量

模拟考试 仿真模拟

题型练习 按题型分类练习

易错题 精选高频易错题

学习资料 考试学习相关信息

搜索

题库预览

A data scientist is working on a deep learning project using cloud-based GPU instances. The dataset consists of millions of small JSON files (each less than 1MB) stored in an object storage service. The data scientist notices that the GPU-accelerated processing is unexpectedly slow. Which of the following approaches would most likely improve performance?

When optimizing hyperparameters in machine learning, what is a key advantage of using Random Search over Grid Search?

After analyzing a large dataset, a data scientist decides to upscale their data processing pipeline to handle increased workloads efficiently. Which cloud platform feature should they prioritize to ensure optimal performance and availability?

What is a key difference between Polars and Dask for distributed data processing?

A data scientist has a dataset with numerical features in ranges of 0-255, 0-65,535, and 0-4,294,967,295. The goal is to optimize memory usage on a GPU- accelerated workflow. Which approach is most appropriate for choosing data types?

Which of the following is the primary benefit of using columnar data format for various data science tasks?

With respect to cuDF spilling in Dask, what happens immediately after a GPU's memory is exhausted?

A data scientist is working on an exploratory data analysis (EDA) project involving a large dataset of 50GB. They need to perform various operations including data loading, filtering, aggregation, and visualization. Which combination of RAPIDS libraries would be most suitable for accelerating this EDA workflow on a GPU?

When training deep learning models with multiple epochs on GPUs, which storage solution typically offers the best performance for large datasets?

When working with GPU-accelerated frameworks like cuDF, which data type is most suitable for a feature representing continuous numerical values in a machine learning model to optimize performance and memory usage?

You are tasked with optimizing an MLOps pipeline involving large datasets and GPU-accelerated workflows. The pipeline currently suffers from slow data loading times. Which combination of techniques would provide the most significant improvement for data throughput in a Dask-cuDF environment?

A machine learning engineer is preparing features such as age and income for model training. After investigation, they noticed that these features have vastly different scales, which could negatively impact the model. They apply a technique to address this issue. Which data preparation task is being performed?

A data scientist at a climate research institute is tasked with processing and analyzing a massive dataset of global weather patterns. The dataset contains hourly measurements from thousands of weather stations worldwide, including temperature, humidity, wind speed, and atmospheric pressure. The data spans several decades and amounts to several terabytes. The goal is to cleanse, preprocess, and standardize this data for use in a complex climate model. Given the scale and nature of the data, which of the following approaches would be most effective?

How does accelerating machine learning workflows help mitigate underfitting and overfitting issues? Pick the 2 correct responses below

Which of the following is not a direct benefit of accelerated data loading?

When designing a distributed ETL pipeline using Dask and cuDF, which of the following optimizations would most significantly reduce memory overhead and improve performance?

When optimizing hyperparameters in machine learning, what is a key advantage of using Random Search over Grid Search?

Which of the following does not benefit from accelerated computing when preparing data for downstream data science tasks?

Which of the following is not a direct benefit of accelerated data loading?

In a data science project, which practices could be used to effectively manage software dependencies? Pick the 2 correct responses below

更多题库