Final Exam: AI Practitioner

WHAT YOU WILL LEARN

• 

  Compare ai practitioner to ai developer and list fundamental differences in their workflows
  

  compare ai practitioner to ml engineer and list fundamental differences in their workflows
  

  identify key benefits of ai optimization and specify improvements which can be achieved from ai optimization
  

  compare ai practitioner to ai engineer and list fundamental differences in their workflows
  

  describe the principle of gradient descent optimization in ai and specify cases in which gradient descent optimization is used
  

  specify the types of ai optimization and describe key differences in the approaches
  

  describe the principle of adagrad optimization in ai and specify cases in which adagrad optimization is used
  

  specify how to tune hyperparameters using grid search approach
  

  work with keras to create and train a feed-forward neural network model and demonstrate its performance
  

  describe how to load and use external data with microsoft cntk
  

  list model types present in amazon ml and specify their purposes
  

  specify how to tune hyperparameters using the bayesian method
  

  describe the principle of stochastic gradient descent optimization in ai and specify cases in which sgd is used
  

  work with python to apply pre-processing techniques to housing price data and troubleshoot cntk machine learning regression model creation and training using this data
  

  describe the main features of intelligent systems and define the concept of iis
  

  define epochs and batch size in cntk and specify how to choose optimal values for best performance
  

  describe the role of hyper parameters in deep learning neural network models
  

  describe the capabilities of amazon ml regarding feature processing
  

  work with python libraries to design an environment for markov decision process for self-driving technology
  

  name primary components of intelligent information systems and their purpose
  

  specify how spark ml pipeline can be used for creating and tuning ml models
  

  list possible challenges and common problems when developing iis
  

  name the features of spark data frame and list useful operations for working with spark data frames
  

  define core and convolutional layers specifying their role in the overall neural network
  

  create training data using spark toolkit and develop spark estimator in python
  

  describe how to create more complex ai models using keras functional api
  

  work with python libraries to build high-level components of markov decision process for self-driving technology
  

  compare and contrast the use of amazon ml and google cloud platform
  

  describe the process of hyperparameter tuning and name multiple approaches to the process
  

  specify cases in which it is advantageous to use amazon ml over other platforms
• 

  identify possible data sources for working with amazon ml
  

  describe keras sequential model api and specify how it is used for developing ai
  

  describe the process of batch prediction in amazon ml and identify cases in which batch prediction is more desirable than online prediction
  

  describe the role of ai practitioner in a company and identify key responsibilities
  

  list possible applications of intelligent information systems
  

  specify multiple approaches to how data can be split using amazon ml
  

  specify the skillset needed to become an ai practitioner and name commonly used tools
  

  list possible operations with resilient distributed datasets and specify their role
  

  work with cntk evaluation tools to evaluate previously created cntk machine learning model
  

  identify how cntk can be used for model visualization
  

  describe the principle of momentum optimization in ai and specify cases in which momentum optimization is used
  

  describe the principle of adam optimization in ai and specify cases in which adam optimization is used
  

  describe the role of hyperparameters in common machine learning models and approaches
  

  work with cntk to create and train a feed-forward neural network as well as demonstrate its performance
  

  define pooling and recurrent layers specifying their role in the overall neural network
  

  list possible sources of data for a spark data frame and describe how to import these into spark
  

  specify the role of ai practitioner when developing ai products and relationship with other developers
  

  specify cases in which it is advantageous to use spark over other platforms
  

  compare and contrast the use of amazon ml and azure ml
  

  recognize why iis development is a growing field and specify demand for iis development
  

  specify multiple techniques and approaches to pre-processing provided by keras
  

  describe how to create a resilient distributed dataset
  

  describe how to create a spark data frame
  

  compare ai practitioner to data scientist/ai scientist and list fundamental differences in their workflows
  

  compare and contrast keras with ms cntk
  

  name multiple libraries which allow for hyperparameter tuning and describe how to use these libraries
  

  describe how real-time prediction is made in amazon ml
  

  specify cases in which it is advantageous to use keras over other platforms
  

  describe the role of hyperparameters in ai development and specify their importance
  

  specify cases in which it is advantageous to use cntk over other platforms