You've developed a model utilizing BigQuery ML for linear regression and aim to retrain it weekly with the cumulative data while minimizing both development effort and scheduling costs. What approach should you take?
#62
You are training an LSTM-based model on Google Cloud AI Platform to summarize text. The job submission script is as follows: bash gcloud ai-platform jobs submit training $JOB_NAME \ --package-path $TRAINER_PACKAGE_PATH \ --module-name $MAIN_TRAINER_MODULE \ --job-dir $JOB_DIR \ --region $REGION \ --scale-tier basic \ -- \ --epochs 20 \ --batch_size=32 \ --learning_rate=0.001 \ You want to ensure that training time is minimized without significantly compromising the accuracy of your model. What should you do?
#63
Your team is developing an application for a global bank, expected to be used by millions of customers. As part of this project, you've built a forecasting model that predicts customers' account balances three days into the future. The goal is to use these predictions to implement a new feature that will notify users when their account balance is likely to fall below $25. How should you deploy and serve these predictions?
#64
You are developing a TensorFlow Extended (TFX) pipeline with standard TFX components. The pipeline includes data preprocessing steps. After deploying the pipeline to production, it will process up to 100 TB of data stored in BigQuery. You need the data preprocessing steps to scale efficiently, publish metrics and parameters to Vertex AI Experiments, and track artifacts using Vertex ML Metadata. How should you configure the pipeline run?
#65
As the lead ML Engineer for your company, you are responsible for building ML models to digitize scanned customer forms. You have developed a TensorFlow model that converts the scanned images into text and stores them in Cloud Storage. You need to use your ML model on the aggregated data collected at the end of each day with minimal manual intervention. What should you do?
#66
You are developing a binary classification ML algorithm that aims to identify whether an image of a scanned document contains a company's logo. However, in the dataset, a significant imbalance exists, with 96% of examples not featuring the logo. To ensure the highest confidence in your model's performance, which metrics should you prioritize?
#67
You are in the process of creating a TensorFlow model for a financial institution, which aims to predict the influence of consumer spending on global inflation. Given the large dataset and the need for extended training, with regular checkpoints, your organization has emphasized cost minimization. What hardware should you select for this task?
#68
You are developing a predictive maintenance model to proactively detect part defects in bridges, and you plan to utilize high-definition bridge images as inputs for your model. To effectively explain the model's output to the relevant stakeholders and enable them to take appropriate action, which approach should you use when building the model?
#69
You require the creation of classification workflows for multiple structured datasets that are currently housed in BigQuery. Since you will need to perform this classification process repeatedly, you aim to execute the following tasks without the need for manual coding: exploratory data analysis, feature selection, model construction, training, hyperparameter tuning, and deployment. What course of action should you take to achieve this?
#70
You hold the role of an ML engineer within a mobile gaming company. A fellow data scientist on your team has recently trained a TensorFlow model, and it falls upon you to integrate this model into a mobile application. However, you've encountered an issue where the current model's inference latency exceeds acceptable production standards. To rectify this, you aim to decrease the inference time by 50%, and you are open to a slight reduction in model accuracy to meet the latency requirement without initiating a new training process. In pursuit of this objective, what initial model optimization technique should you consider for latency reduction?
#71
You trained a model, packaged it with a custom Docker container for serving, and deployed it to Vertex AI Model Registry. When you submit a batch prediction job, it fails with this error: "Error model server never became ready. Please validate that your model file or container configuration are valid." There are no additional errors in the logs. What should you do?
#72
You work for a pet food company that manages an online forum. Customers upload photos of their pets on the forum to share with others. About 20 photos are uploaded daily. You want to automatically and in near real-time detect whether each uploaded photo has an animal. You want to prioritize time and minimize the cost of your application development and deployment. What should you do?
#73
You have established an ML pipeline featuring various input parameters, and your objective is to explore the trade-offs among different combinations of these parameters. The parameters in question include: - The input dataset - The maximum tree depth for the boosted tree regressor - The learning rate for the optimizer You need to assess the pipeline's performance for the various parameter combinations, evaluating them in terms of F1 score, training time, and model complexity. It is essential for your methodology to be reproducible, and you aim to track all runs of the pipeline on a consistent platform. What steps should you take to achieve this?
#74
You've received a dataset containing sales predictions derived from your company's marketing efforts. This well-structured data is stored in BigQuery and has been meticulously maintained by a team of data analysts. Your task is to create a report that offers insights into the predictive potential of the data. You've been instructed to run various ML models, ranging from basic models to complex multilayered neural networks. You have only a limited amount of time to collect the results of your experiments. Which Google Cloud tools should you employ to efficiently and independently accomplish this task?
#75
You are implementing transfer learning to train an image classifier, leveraging a pre-trained EfficientNet model. Your training dataset consists of 20,000 images, and your intention is to retrain the model on a daily basis. To keep infrastructure costs to a minimum, what platform components and configuration environment should you employ?
#76
Your data science team is tasked with conducting rapid experiments involving various features, model architectures, and hyperparameters. They need an efficient way to track the accuracy metrics of these experiments and access the metrics programmatically over time. What approach should they take to achieve this while minimizing manual effort?
#77
You work for a company that captures live video footage of checkout areas in their retail stores. Your task is to build a model to detect the number of customers waiting for service in near real-time. You aim to create this solution quickly and with minimal effort. What approach should you take to build the model?
#78
Your team is in the process of developing a convolutional neural network (CNN)-based architecture from the ground up. Initial experiments conducted on your on-premises CPU-only infrastructure have shown promising results, but the model's convergence is slow. To expedite the model training process and shorten time-to-market, you are considering conducting experiments on Google Cloud virtual machines (VMs) equipped with more powerful hardware. It's important to note that your code doesn't involve manual device placement, and it hasn't been encapsulated within the Estimator model-level abstraction. Given this context, which environment should you choose for training your model?
#79
You are employed at a bank and have a custom tabular ML model provided by the bank's vendor. Unfortunately, the training data for this model is sensitive and unavailable. The model is packaged as a Vertex AI Model serving container, and it accepts a string as input for each prediction instance. Within these strings, feature values are separated by commas. Your objective is to deploy this model into production for online predictions while also monitoring the feature distribution over time with minimal effort. What steps should you take to achieve this?
#80
Your data science team has requested a system that supports scheduled model retraining, Docker containers, and a service that supports autoscaling and monitoring for online prediction requests. Which platform components should you select for building this system?
