Daily Quiz ( 23.12.2020 )

Which is not the application of AI ?

• Face Recognition

• Agricultural Automation

• DevOps

• Customer Support Automation

• Sentiment Analysis

What is the correct approach of AI to start with the data?

• Data ingestion → Data Cleaning & Transformation → Model Training → Testing & Validation → Deployment

• Data Cleaning → Transformation → Data Ingestion → Testing & Validation → Model Training → Deployment

• Data Cleaning & Transformation → Testing & Validation → Data Ingestion → Model Training → Deployment

• Data ingestion → Data Cleaning & Transformation → Model Training → Testing & Validation → Model Selection → Model Training → Testing & Validation → Deployment

Which step is the most significant in AI approach at production level?

• Deployment

• Model Building

• Business Problem

• Data

• Active Learning & Tuning

What is it called when a model learns from data that is labeled by human annotators and experts?

• Transfer Learning

• Passive Learning

• Active Learning

When a model does not know the answer to a specific query or how it might respond to a certain input, what you should do?

• Fine tune your algorithm

• Try different models as trial and error and get the best optimized solution

• Put more effort on data preprocessing

• More data from human annotators and re-train model

In which case(s) would you consider updating a classification model? (Check ALL that apply.)

• The model accuracy has been degrading over time

• The model accuracy has stayed the same over time

• The underlying data changes and a new class is introduced

• All of the above

_____ owns human-computer interaction design and ____ ensures the entire pipeline can support rapid development and iteration after launch.

• DevOps engineer, Business Owner

• Business Owner, Designer

• Designer, Data Engineer

• Designer, Software Engineer

Scalability testing and Functional testing are done by ___

• DevOps Engineer

• Designer

• Quality Assurance Engineer

• Data Scientist

What causes a higher error rate in speech recognition?

• noisy environment

• accented speech

• gender of the speaker

• limited training data

Which of the following can predict outcomes from unstructured inputs?

• machine learning

• deep learning

• artificial intelligence

• all of the above