Quick Notes - DOMAIN 1: AWS Certified AI Practitioner

If you are prepping for the AWS Certified AI Practitioner https://aws.amazon.com/certification/certified-ai-practitioner/, these notes should be enough to get the fundamentals for the exam.

Domain 1: Fundamentals of AI and ML

• Building a machine learning model involves data collection and preparation, selecting an appropriate algorithm, training the model on the prepared data, and evaluating its performance through testing and iteration.

Training Data -> ML Algorithm -> Model (create a picture)

Training Data:

There are a several different types of data used in training an ML model. First, it's important to know the difference between labeled and unlabeled data.

• Labeled Data - Data with labels (classification)

• Unlabeled Data - Images without any tags/labels

The main types of data used in training are structured and unstructured data. They come with various subtypes, which you can find by expanding the following categories.

• Structured data - spreadsheets, databases, or CSV files,

• UnStructured data - digital images, photographs, and video frames.

Machine learning:

The compiled training data is fed into machine learning algorithms. The ML learning process is traditionally divided into three broad categories:

• supervised learning - the algorithms are trained on labeled data.

• unsupervised learning - algorithms that learn from unlabeled data.

• reinforcement learning - the machine is given only a performance score as guidance and semi-supervised learning, where only a portion of training data is labeled.

Inferencing:

After the model has been trained, it is time to begin the process of using the information that a model has learned to make predictions or decisions. This is called inferencing.

Batch inferencing: Two main types of Batch inferencing is when the computer takes a large amount of data, such as images or text, and analyzes it all at once to provide a set of results. This type of inferencing is often used for tasks like data analysis, where the speed of the decision-making process is not as crucial as the accuracy of the results in machine learning.

Real-time inferencing: Real-time inferencing is when the computer has to make decisions quickly, in response to new information as it comes in. This is important for applications where immediate decision-making is critical, such as in chatbots or self-driving cars. The computer has to process the incoming data and make a decision almost instantaneously, without taking the time to analyze a large dataset.

Deep Learning Fundamentals

Neural networks: At the core of deep learning are neural networks. Just like our brains have neurons that are connected to each other, neural networks have lots of tiny units called nodes that are connected together. These nodes are organized into layers. The layers include an input layer, one or more hidden layers, and an output layer.

The following are a couple branches of AI where deep learning is used to enhance results.

• Computer vision is a field of artificial intelligence that makes it possible for computers to interpret and understand digital images and videos. Deep learning has revolutionized computer vision by providing powerful techniques for tasks such as image classification, object detection, and image segmentation.
  

• Natural language processing (NLP) is a branch of artificial intelligence that deals with the interaction between computers and human languages. Deep learning has made significant strides in NLP, making possible tasks such as text classification, sentiment analysis, machine translation, and language generation.

Generative AI Fundamentals

Generative AI is powered by models that are pretrained on internet-scale data, and these models are called foundation models (FMs). With FMs, instead of gathering labeled data for each model and training multiple models as in traditional ML, you can adapt a single FM to perform multiple tasks. These tasks include text generation, text summarization, information extraction, image generation, chatbot interactions, and question answering. FMs can also serve as the starting point for developing more specialized models.

FM lifecycle:

Data Selection -> Pre-training -> Optimization (Pre-trained language models can be optimized through techniques like prompt engineering, retrieval-augmented generation (RAG), and fine-tuning on task-specific data.) -> Evaluation (measured using appropriate metrics and benchmarks.) -> Deployment -> Feedback and continuous improvement

Amazon Bedrock provides access to a choice of high-performing FMs from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI, and Amazon.

With these FMs as a foundation, you can further optimize their outputs with prompt engineering, fine-tuning, or RAG.

There are a few types of FMs that are essential to understanding generative AI's capabilities.

LLM: Large language models (LLMs) can be based on a variety of architectures, but the most common architecture in today's state-of-the-art models is the transformer architecture.

To better understand how LLMs work:

• tokens: Tokens are the basic units of text that the model processes.

• embeddings and vectors: Embeddings are numerical representations of tokens, where each token is assigned a vector (a list of numbers) that captures its meaning and relationships with other tokens. These vectors are learned during the training process and allow the model to understand the context and nuances of language.

Diffusion models: Diffusion is a deep learning architecture system that starts with pure noise or random data. Diffusion models learn through a two-step process of forward diffusion and reverse diffusion. (text-to-image models)

Multimodal models: Process and generate multiple modes of data simultaneously.

Optimizing model outputs

An FM can be further optimized in several different ways:

• Prompt engineering: Prompt engineering focuses on developing, designing, and optimizing prompts to enhance the output of FMs for your needs.

• Fine-tuning: fine-tuning will change model weights.

  • Prompt tuning is a type of instruction fine-tuning.

  • Reinforcement learning from human feedback (RLHF) provides human feedback data, resulting in a model that is better aligned with human preferences.

• Retrieval-augmented generation(RAG): is a technique that supplies domain-relevant data as context to produce responses based on that data.

AWS AI/ML Services:

• In the text and documents domain, Amazon Comprehend for natural language processing, data insights.

• Amazon Translate for language translation.

• Amazon Textract for extracting data from scanned documents.

• For chatbots, AWS offers Amazon Lex, which you can use to build conversational interfaces powered by the same deep learning technologies that drive Amazon Alexa.
  

• In the speech domain, you can find Amazon Polly for text-to-speech and

Amazon Transcribe for automatic speech recognition.

• In the vision domain, you have Amazon Rekognition, a deep learning-based computer vision service that can analyze images and videos for a wide range of applications.

• For search, Amazon Kendra reimagines enterprise search for websites and applications so that individuals can readily find the content they are looking for.

• In the recommendations domain, we have Amazon Personalize for real-time personalization and recommendations.

Amazon SageMaker JumpStart, which provides a set of solutions for the most common use cases.

If you'd prefer to experiment with building AI applications, you can get hands-on experience by using PartyRock, an Amazon Bedrock Playground.

Amazon Q, a generative AI–powered assistant designed for work that can be tailored for a business's data.

Amazon Q Developer, providing ML–powered code recommendations to accelerate development in a variety of programming languages and applications.

For Domain 2 -> https://www.bansalonsecurity.com/post/quick-notes-domain-2-aws-certified-ai-practitioner

Reference: AWS SkillBuilder https://skillbuilder.aws/learn