Image Recognition

What image recognition technique is used when using large sets of data?

• Deep learning

• Machine Learning

• Neural Networks

• Faster RCNN

What is the first step to building an image recognition model?

• preprocessing images

• building a data pipeline

• partitioning the data set

• selecting a model

Can this image be used to train the model?

• Yes

• No

Why is cleaning the dataset such an important step?

• All data is viable

• It's easier to organize the data

• Clean data increases efficiency

• corrupt data can lead to inaccuracies

True or False: Testing data should be completely new data

• False

• True

What is back propagation used for?

• calculating weight and bias values

• sending output from the previous layer as input to the next layer

• calculating the gradient of error

• predicting the output

How can you counter overfitting data?

• Use data that is specific to the training model

• Use plenty of training and validation data

• Reuse training data as validation data

• Rebuild the model with new data

What is the ratio of training data to validation data?

• 80:20

• 20:80

• 50:50

• 64:36

What type of model is depicted in this diagram?

• Supervised learning

• Unsupervised learning

• Reinforcement learning

• Neural Networks

Which of the following is NOT a feature of pretrained models?

• Better speed, accuracy, and size

• Trained on large data sets and can be fine tuned to a specific set

• Provides an initial set of weights and biases

• Built from scratch