book

AI-Assisted Programming for Web and Machine Learning

Name: AI-Assisted Programming for Web and Machine Learning
ISBN: 9781835086056

by Christoffer Noring, Anjali Jain, Marina Fernandez, Ayşe Mutlu, Ajit Jaokar

August 2024

Intermediate to advanced

602 pages

13h 29m

English

Packt Publishing

Read now

Unlock full access

Preface
Who this book is forWhat this book coversTo get the most out of this bookGet in touch
It’s a New World, One with AI Assistants, and You’re Invited
IntroductionHow ChatGPT came to be, from NLP to LLMsThe rise of LLMsGPT modelsHow LLMs are betterThe new paradigm, programming with natural languageChallenges and limitationsAbout this bookWho this book is forEvolution of programming languagesLooking aheadHow to use this book
Prompt Strategy
IntroductionWhere you areGuidelines for how to prompt efficientlyPrompt techniquesTask-Action-Guideline prompt pattern (TAG)Persona-Instruction-Context prompt pattern (PIC)Exploratory prompt patternLearn-Improvise-Feedback-Evaluate prompt pattern (LIFE)Which pattern to choose?Prompt strategy for web developmentBreak down the problem: “web system for inventory management”Further breakdown of the frontend into featuresGenerate prompts for each featureIdentify some basic principles for web development, a “prompt strategy”Prompt strategy for data scienceProblem breakdown: predict salesFurther breakdown into features/steps for data scienceGenerate prompts for each stepIdentify some basic principles for data science, “a prompt strategy for data science”Validate the solutionVerification via promptsClassical verificationSummary
Tools of the Trade: Introducing Our AI Assistants
IntroductionUnderstanding CopilotHow Copilot knows what to generateCopilot capabilities and limitsSetup and installationInstalling CopilotGetting started with CopilotAssignment: improve the codeSolutionChallengeReferencesUnderstanding ChatGPTHow does ChatGPT work?ChatGPT capabilities and limitsSetup and installationGetting started with ChatGPTPromptingSummary
Build the Appearance of Our App with HTML and Copilot
IntroductionBusiness problem: e-commerceProblem domainProblem breakdown: identify the featuresPrompt strategyPage structureAdd AI assistance to our page structureYour first prompt, simple prompting, and aiding your AI assistantYour second prompt: adding more contextYour third prompt: accept prompt suggestionsChallenge: vary the promptUse case: build a front for an e-commerceLogin pageProduct list pageRemaining pagesAssignmentChallengeQuizSummary
Style the App with CSS and Copilot
IntroductionBusiness problem: e-commerceProblem and data domainBreaking the problem down into featuresPrompting strategyCSS, or Cascading Style SheetsFirst CSSCSS by nameAssignmentSolutionUse case: style the e-commerce appBasket pageChallengeQuizSummary
Add Behavior with JavaScript
IntroductionBusiness problem: e-commerceProblem and data domainBreaking the problem down into featuresPrompting strategyAdding JavaScriptThe role of JavaScriptAdding JavaScript to a pageA second example: adding a JavaScript library/frameworkChallengeUse case: adding behaviorImproving the outputAdding BootstrapAdding Vue.jsAssignmentSolutionSummary
Support Multiple Viewports Using Responsive Web Layouts
IntroductionBusiness problem: e-commerceProblem and data domainBreaking the problem down into featuresPrompting strategyViewportsMedia queriesWhen to adjust to different viewports and make it responsiveUse case: make our product gallery responsiveAssignmentSolutionChallengeSummary
Build a Backend with Web APIs
IntroductionBusiness domain: e-commerceProblem and data domainFeature breakdownPrompt strategyWeb APIsWhat language and framework should you pick?Planning the Web APICreating a Web API with Python and FlaskStep 1: Create a new projectStep 2: Install FlaskStep 3: Create an entry pointStep 4: Create a Flask appUse case: a Web API for an e-commerce siteStep 1: Create a Web API for an e-commerce siteStep 2: Return JSON instead of textStep 3: Add code to read and write to a databaseStep 4: Improve the codeRun the codeRefactor the codeStep 5: Document the APIAssignmentSolutionChallengeSummary
Augment Web Apps with AI Services
IntroductionBusiness domain, e-commerceProblem and data domainFeature breakdownPrompt strategyCreating a modelComing up with a planImporting librariesReading the CSV fileCreating test and training datasetsCreating a modelHow good is the model?PredictSaving the model to a .pkl fileCreating a REST API in PythonConverting the model to ONNXCreating a model in ONNX formatLoading the ONNX model in JavaScriptInstalling onnxruntime in JavaScriptLoading the ONNX model in JavaScriptAssignment: Build a REST API in JavaScript that consumes the modelSolutionQuizSummary

Maintaining Existing Codebases
IntroductionPrompt strategyDifferent types of maintenanceThe maintenance processAddressing a bug1. Identify the problem2. Implement the changeAdding a new feature 1. Identify a problem and find the function/s to change2. Implement change, and add a new feature and testsImproving performanceBig O notation calculationMeasuring performanceImproving maintainability1. Identify the problems. What problems do you see?2. Add tests and de-risk change3. Implement change and improve maintainabilityChallengeUpdating an existing e-commerce siteAssignmentKnowledge checkSummary
Data Exploration with ChatGPT
IntroductionBusiness problemProblem and data domainDataset overviewFeature breakdownPrompting strategyStrategy 1: Task-Actions-Guidelines (TAG) prompt strategyStrategy 2: Persona-Instructions-Context (PIC) prompt strategyStrategy 3: Learn-Improvise-Feedback-Evaluate (LIFE) prompt strategyData exploration of the Amazon review dataset using the free version of ChatGPTFeature 1: Loading the datasetFeature 2: Inspecting the dataFeature 3: Summary statisticsFeature 4: Exploring categorical variablesFeature 5: Rating distributionFeature 6: Temporal trendsFeature 7: Review length analysisFeature 8: Correlation studyData exploration of the Amazon review dataset using ChatGPT-4oAssignmentChallengeSummary
Building a Classification Model with ChatGPT
IntroductionBusiness problemProblem and data domainDataset overviewBreaking the problem down into featuresPrompting strategyStrategy 1: Task-Actions-Guidelines (TAG) prompt strategyStrategy 2: Persona-Instructions-Context (PIC) prompt strategyStrategy 3: Learn-Improvise-Feedback-Evaluate (LIFE) prompt strategyBuilding a sentiment analysis model to accurately classify Amazon reviews using the free version of ChatGPTFeature 1: Data preprocessing and feature engineeringFeature 2: Model selection and baseline trainingFeature 3: Model evaluation and interpretationFeature 4: Handling imbalanced dataFeature 5: Hyperparameter tuningFeature 6: Experimenting with feature representationBuilding a sentiment analysis model to accurately classify Amazon reviews using ChatGPT-4 or ChatGPT PlusFeature 1: Data preprocessing and feature engineeringFeature 2: Model selection and baseline trainingFeature 3: Model evaluation and interpretationFeature 4: Handling data imbalanceFeature 5: Hyperparameter tuningFeature 6: Experimenting with feature representationAssignmentChallengeSummary
Building a Regression Model for Customer Spend with ChatGPT
IntroductionBusiness problemProblem and data domain Dataset overviewBreaking the problem down into featuresPrompting strategyStrategy 1: Task-Actions-Guidelines (TAG) prompt strategyStrategy 2: Persona-Instructions-Context (PIC) prompt strategyStrategy 3: Learn-Improvise-Feedback-Evaluate (LIFE) prompt strategy Building a simple linear regression model to predict the “Yearly Amount Spent” by customers using the free version of ChatGPTFeature 1: Building the model step by stepFeature 2: Applying regularization techniquesFeature 3: Generating a synthetic dataset to add complexityFeature 4: Generating code to develop a model in a single step for a synthetic datasetLearning simple linear regression using ChatGPT PlusFeature 1: Building a simple linear regression model step by stepFeature 2: Applying regularization techniquesFeature 3: Generating a synthetic dataset to add complexityFeature 4: Generating code to develop a model in a single step for a synthetic datasetAssignmentChallengeSummary
Building an MLP Model for Fashion-MNIST with ChatGPT
IntroductionBusiness problemProblem and data domainDataset overviewBreaking the problem down into featuresPrompting strategyStrategy 1: Task-Actions-Guidelines (TAG) prompt strategyStrategy 2: Persona-Instructions-Context (PIC) prompt strategyStrategy 3: Learn-Improvise-Feedback-Evaluate (LIFE) prompt strategyBuilding an MLP model to accurately classify the Fashion-MNIST images using the free version of ChatGPTFeature 1: Building the baseline modelFeature 2: Adding layers to the modelFeature 3: Experimenting with batch sizesFeature 4: Experimenting with the number of neuronsFeature 5: Trying different optimizersAssignmentChallengeSummary
Building a CNN Model for CIFAR-10 with ChatGPT
IntroductionBusiness problemProblem and data domainDataset overviewBreaking the problem down into featuresPrompting strategyStrategy 1: Task-Actions-Guidelines (TAG) prompt strategyStrategy 2: Persona-Instructions-Context (PIC) prompt strategyStrategy 3: Learn-Improvise-Feedback-Evaluate (LIFE) prompt strategyBuilding a CNN model to accurately classify the CIFAR-10 images using the free version of ChatGPTFeature 1: Building the baseline CNN model with a single convolutional layerFeature 2: Experimenting with the addition of convolutional layersFeature 3: Incorporating dropout regularizationFeature 4: Implementing batch normalizationFeature 5: Optimizing with different optimizersFeature 6: Applying the DavidNet architectureAssignmentChallengeSummary
Unsupervised Learning: Clustering and PCA
IntroductionBreaking the problem down into featuresPrompt strategyCustomer segmentation The datasetAdding AI assistance to the unsupervised learning model development process Load the dataset Inspect the data Summary statistics Preprocessing the data Feature engineering Checking for outliersRemoving outliersData scaling using standardizationDeciding on the number of clusters Creating a clustering modelVisualize clustering resultsFinal thoughts on clustering and the prompting processProduct clustering for an e-commerce projectYour initial prompt: Set context Load and preprocess the dataFeature engineering and text data preprocessingFeature engineeringChoose clustering algorithm Feature scaling Apply clustering algorithm Interpret clusters and visualize results Interpreting clusterVisualizing clustersCreating a word cloudAssigning categories to products and evaluating and refiningEvaluate and refineReflection on prompts for this use case AssignmentSolutionSummary
Machine Learning with Copilot
IntroductionGitHub Copilot Chat in your IDEHow it worksDataset overviewSteps for data explorationPrompt strategyYour initial data exploration prompt: Prompt 1, setting the high-level contextStep 1: Load the dataset Running the code for loading dataStep 2: Inspect the data Step 3: Summary statisticsStep 4: Explore categorical variablesStep 5: Distribution of ratings Step 6: Temporal analysisStep 7: Review length analysisStep 8: Correlation analysis Step 9: Additional exploratory analysisStep 10: Text PreprocessingStep 11: Word Frequency AnalysisStep 12: Sentiment Score CalculationText preprocessingWord frequency analysisSentiment score calculationStep 13: Visualize the Distribution of Sentiment ScoresStep 14: Analyze the Relationship Between Sentiment Score and Other VariablesVisualize the distribution of sentiment scoresAnalyze the relationship between sentiment score and other variablesAssignmentSolutionSummary
Regression with Copilot Chat
IntroductionRegressionDataset overview Explore the dataset Prompt strategyYour initial prompt Exploratory data analysisData splitting Build a regression model Evaluate the model Evaluation metrics AssignmentSummary
Regression with Copilot Suggestions
IntroductionDataset overview Prompt strategyStart coding with Copilot’s help Step 1: Import libraries with Copilot’s assistance Step 2: Load and explore the datasetGet types and columnsShape of the datasetAddressing the column typesStatistical summaryCheck for missing valuesCheck for duplicatesScale numerical featuresVisualizationStep 3: Split data into training and testing sets Asking questionsStep 4: Build a regression problem Step 5: Train the model Step 6: Evaluate model performance AssignmentSummary
Increasing Efficiency with GitHub Copilot
IntroductionCode generation and automationCopilot’s active editorCopilot ChatCopilot commandsCreating a NotebookCreating a projectDebugging and troubleshootingCode review and optimization techniquesWorkspaceVisual Studio Code lookupTerminalAssignmentChallengeQuizSummary
Agents in Software Development
IntroductionWhat are agents?How do agents work?Simpler agents versus agents using AISimpler agentsA simple agent is not a great conversationalistImproved conversation with tool calling and large language models (LLMs)The anatomy of a conversational agentMore on tool calling in LLMsAdding capabilities to GPT using toolsAdvanced conversationsModeling advanced conversationsPseudo code for advanced conversationsAutonomous agentsAssignmentChallengeQuizSummaryReferences
Conclusion
Recap of the bookMajor conclusionsWhat’s nextAt last
Other Books You May Enjoy
Index

Content preview from AI-Assisted Programming for Web and Machine Learning

3 Tools of the Trade: Introducing Our AI Assistants

Introduction

Writing code or text takes time, at least if you want it to be well organized and readable. But what if you could have a tool that would help you write code faster and with less effort? That’s what GitHub Copilot and ChatGPT are really all about.

Before you start using an AI assistant, it’s a good idea to get a mile-high view of its capabilities and limitations. You want to see what you can and can’t use it for, or at least understand where the tool performs less well.

In this chapter, we will cover the following:

Understanding what GitHub Copilot and ChatGPT are and how they work
Learning about Copilot’s capabilities and limits
Installing GitHub Copilot
Generating code completions ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Python Machine Learning By Example - Fourth Edition

Publisher Resources

ISBN: 9781835086056

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills

AI-Assisted Programming for Web and Machine Learning

by Christoffer Noring, Anjali Jain, Marina Fernandez, Ayşe Mutlu, Ajit Jaokar

3

Tools of the Trade: Introducing Our AI Assistants

Introduction

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.