Skip to content

CAIML.md

Latest commit

 

History

History
67 lines (58 loc) · 5.58 KB

CAIML.md

File metadata and controls

67 lines (58 loc) · 5.58 KB

Cloud-based Artificial Intelligence and Machine Learning - CAIML

Simply stated, a cloud-based artificial intelligence (AI) solution is about providing relevant infrastructure and machines that can perform tasks that are characteristic of human intelligence. The massive data storage and compute ability provided by the cloud has increased the ability to perform machine learning (ML) and deep learning (DL). While AI encompasses a wide range of approaches and solutions about mimicking human intelligence, ML is a way of achieving AI, by relying on (cloud-based) machines, and related data sources. DL is one of the many approaches to ML, that mimics the biological structure of the brain, and involves using novel techniques, such as artificial neural networks in combination with utilizing large data sets.

In this KA, students should be able to observe, identify, and explain core cloud mechanisms that help build relevant AI solutions. They should also explain the difference between using cloud-hosted AI/ML solutions that many cloud providers offer, and working on a do-it-yourself solution that utilizes existing core resources (compute, storage, network, etc.) on the cloud infrastructure. Students - based on their understanding level and focus on the subject matter

  • should be able to outline a plethora of relevant ML/AI tools and methodologies, and narrow it down to pick a relevant experimental methodology necessary to perform statistical analysis of large-scale data sources.

Each of the following Learning Objectives links to a list of materials that can be used to teach the LO.

Conceptual Learning Objectives

  • CAIML-CL1: Appreciate the distinction between the popular view of the field and the actual research results, and the fact that the computational complexity of most AI problems requires us to deal with approximation techniques regularly.

  • CAIML-CL2: Explore the processing hardware technology (CPU, GPU, TPU, and so on) used by popular ML/DL solutions.

  • CAIML-CL3: Appreciate resource discovery and self-organization methods used in utilizing unstructured networks.

  • CAIML-CL4: Identify the ethical implications of how an AI application or service assists humanity, and whether it is designed for intelligent privacy and not for deceiving humans.

  • CAIML-CL5: Identify features of common AI workloads.

  • CAIML-CL6: Describe Artificial Intelligence workloads and considerations.

  • CAIML-CL7: Identify guiding principles for responsible AI.

  • CAIML-CL8: Describe fundamental principles of machine learning.

  • CAIML-CL9: Identify common machine learning types.

  • CAIML-CL10: Describe core machine learning concepts.

  • CAIML-CL11: Identify core tasks in creating a machine learning solution.

  • CAIML-CL12: Describe capabilities of no-code machine learning.

  • CAIML-CL13: Describe features of computer vision workloads.

  • CAIML-CL14: Identify common types of computer vision solutions.

  • CAIML-CL15: Identify tools and services for computer vision tasks.

  • CAIML-CL16: Describe features of Natural Language Processing.

  • CAIML-CL17: Identify features of common NLP Workload Scenarios.

  • CAIML-CL18: Identify tools and services for NLP workloads.

  • CAIML-CL19: Describe features of conversational AI workloads.

  • CAIML-CL20: Identify common use cases for conversational AI.

  • CAIML-CL21: Identify services for conversational AI.

  • CAIML-CL22: Understand the data science pipeline of data exploration, preparation, visualization, and insights using cloud services.

  • CAIML-CL23: Highlight the key infrastructure, querying, and cost challenges faced by data analysts when using Cloud Services.

  • CAIML-CL24: Explore building bespoke machine learning models using a Cloud based AI Platform.

  • CAIML-CL25: Discuss machine learning in the cloud.

  • CAIML-CL26: Leverage available cloud services to create custom machine learning models.

Experiental Learning Objectives

  • CAIML-EL1: Use cloud-based scientific computational environments (e.g., Jupyter Notebook) to produce interactive code that interacts with large data sets along with results of running the code and human-readable descriptions.

  • CAIML-EL2: Demonstrate the ability to identify and critique a subset of use cases for learning and inference, such as classification and regression.

  • CAIML-EL3: Demonstrate implementation of underlying algorithms in statistically valid experiments, including the design of baselines, evaluation metrics, statistical testing of results, and provision against over training.

  • CAMIL-EL4: Train, Test and Deploy a Cloud Hosted Machine Learning Model.

  • CAMIL-EL5: Explore cognitive cloud service.s

  • CAMIL-EL6: Apply a range of pre-trained machine learning models using

  • CAMIL-EL7: Analyze and visualize data using Cloud Services.

  • CAMIL-EL8: Employ cloud services APIs for cognitive based services such as vision, speech, and text.