10 Days of No Code Artificial Intelligence Bootcamp

Explore how AI fuels a virtual stylist that classifies fashion categories in Instagram and Facebook images, while you learn to build and deploy AI models with Google Teachable Machines.

Build a no-code AI model to classify fashion items from images using Google Teachable Machines. Demonstrate a project card and a ten-class demo with real-world retailer applications.

Explore AI applications in fashion, including a virtual stylist that analyzes customer Instagram and Facebook images to classify fashion categories, forecast trends, and drive targeted marketing campaigns.

Explore data exploration by building a supervised ai classifier with Google Teachable Machines, training on 1000 fashion images across ten classes, then test and deploy with new images.

Learn to train a fashion image classifier with Google Teachable Machines by uploading 1,000 images across 10 classes, then train and test the model on new images to assess performance.

Learn to export and deploy a trained model from Google Teachable Machines, create an accessible endpoint, and test it with sample images without writing code.

Extend your no-code ai bootcamp project by retraining a model with 13 classes—adding socks, wallets, and scarves—train for 50 epochs, test with new images, and export to deploy cloud model.

Celebrate finishing day one of the no code artificial intelligence bootcamp, acknowledge your milestone and newly acquired skills, and look forward to tomorrow's content plus your feedback for updates.

Explore ai fundamentals and how it learns, including learning rate, epochs, and batch size. Assess model performance with confusion matrix, accuracy, and error rate, and learn to tweak hyperparameters.

AI basics, including an artificial neural network, master key terminologies, and learn how to train and evaluate models using epoch, batch size, learning rate, accuracy, precision, recall, and confusion matrices.

Explore how artificial neural networks mimic brain neurons, including inputs, weights, bias, and activation, to build a trainable model that can classify data.

Understand how to train and test artificial neural networks using supervised learning, including 80/20 data splits, weight optimization, and generalization to unseen data.

Learn core ai lingo by explaining epochs, batch size, and learning rate. Compare training and testing progress using accuracy and loss metrics.

Learn how to read a confusion matrix to assess classifier performance, including true/false positives and negatives, and compute accuracy and misclassification rate with cancer and fashion examples.

Students tackle final project part a, analyzing a 13-class model with confusion matrix metrics, accuracy, and misclassification rate while adjusting epochs, batch size, and learning rate.

Train a five-class sign language image classifier using a no-code platform, evaluate with confusion matrix and accuracy, test on unseen images, and export and deploy the model.

Celebrate overcoming a tough day of technical content and mastering AI fundamentals. Rest up for brand new content tomorrow and share feedback to shape future course launches.

Collect live training and testing data with your webcam to train and deploy a face mask detection model using Google Teachable Machines, tweaking hyperparameters for better indoor accuracy.

Collect live training data with a webcam to train, test, and deploy a no-code ai model that detects mask wearing, and classify four states from no mask to properly worn.

Examine the business case for ai-driven mask detection to enforce public health guidelines. Build a simple ai model with training data and explore facial recognition applications.

Learn to collect live training data with Google Teachable Machines using a webcam, build two classes—no mask detected and mask detected—and optimize lighting and background for better model training.

Train a model in Google Teachable Machines with live webcam data for two classes, then tune epochs, batch size, and learning rate while monitoring training progress.

Deploy and evaluate a trained Google Teachable Machines model, export and upload it, and test its live performance via webcam with no mask and mask detection.

Explore classifier models KPIs, including confusion matrix, accuracy, misclassification rate, precision, and recall; learn how unbalanced data can mislead accuracy and affect true/false positives/negatives.

Explore precision vs. recall through real-world binary classification examples like bank fraud detection and spam filtering, illustrating true positives, false positives, and false negatives via a confusion matrix.

Extend the face mask detector with two extra classes in the final project, collect 300 images per class, tune hyperparameters, analyze the confusion matrix, train, test, and deploy.

Visualize artificial neural networks and architectures like residual networks using TensorFlow Playground, Space JS, and Ryerson 3D convolutional visualizations; learn activation functions, train feedforward nets, and explore transfer learning.

Visualize deep neural networks by exploring layers with tensor space.js, TensorFlow Playground, and Ryerson CNN visualizations, and learn how features are extracted to classify digits 0–9.

Delve into the fundamentals of artificial neural networks, from single neurons to multi-layer feedforward and convolutional networks, and visualize their layers and weights with TensorFlow Playground.

Visualize feedforward neural networks in TensorFlow Playground to explore how activation functions, hidden layers, and learning rate affect class separation and training versus test loss.

Configure a feedforward neural network on the spiral dataset, set an 80/20 split with 30% noise and batch size 20, then tune architecture and hyperparameters for classification.

Explore convolutional neural networks by learning how kernels extract features from images, apply pooling and flattening, and feed them into a dense network, with relu and real-world chest X-ray example.

Explore convolutional neural networks visually using Ryerson 2D and 3D visualizers, showing input digits, convolution layers, downsampling, flattening, dense layers, and the final outputs.

Explore the LeNet convolutional neural network architecture: six convolution layers on 28x28 inputs, subsampling to 14x14 and 5x5, then a 120–84 fully connected classifier producing 10 outputs.

Visualize a net convolutional neural network with feature maps from 28x28 to 14x14 to 10x10 and 5x5, flattening to a dense 84-node output predicting digits 0–9 using tensor space.js.

Explore residual neural networks (resnet) and how skip connections overcome the vanishing gradient to enable hundreds of layers. Visualize the ResNet-50 model in TensorSpace.js and learn its ImageNet performance.

Compare alexnet to resnet, highlighting eight-layer conv nets and skip connections, then visualize alexnet in tensor space dot js and in TensorFlow.js, and complete the final project part B.

Celebrate completing day 4 as you generate ai visualizations. Shift to a brand new tool tomorrow and provide feedback for future course launches.

Learn to use data robot to build, train, test, and deploy no code AI models that predict used car prices from features like make, model, engine size, horsepower, and mpg.

Learn to build, train, test, and deploy a no-code machine learning model in DataRobot to predict car prices from features like horsepower, cylinders, engine size, and make.

Discover real-world ai applications with price prediction stories, including fair price forecasting for flights, real estate recommendations, and market segmentation, and learn how self-learning models guide optimal purchasing decisions.

Investigate data for a regression model by reviewing inputs such as make, model, type, origin, drivetrain, engine size, cylinders, horsepower, mpg city/highway, weight, length, wheelbase, and output MSRP.

Upload data to DataRobot, sign up for a free trial, and create a new project by dragging and dropping the used car price dataset.

Perform exploratory data analysis in DataRobot, select MSRP as the target, and review features such as make, origin, drivetrain, engine size, horsepower, MPG, wheelbase, and weight, using a train-validation-holdout split.

Train and evaluate multiple regression models in DataRobot, conduct feature association analysis, and select the best Keras residual neural network for deployment.

DataRobot analyzes regression model performance using a leaderboard and R squared, comparing multiple models and highlighting horsepower as the key feature behind predictions.

Deploy a top model in DataRobot without writing code, create an application using the predictor, and run inferences to see how horsepower and other features influence car price.

Explore how simple and multiple linear regression differ, how training data predicts vehicle MSRP from engine size and other variables, and how least-squares fit determines the model.

Explore selecting the best deployment model, tune hyperparameters including the learning rate with advanced tuning in DataRobot, train and assess performance against the leaderboard using R^2.

Use the final project insurance data in DataRobot to perform exploratory data analysis, train multiple regression models, deploy the best model, and infer insurance charges.

Reach the halfway milestone in the 10 days of no code artificial intelligence bootcamp. Rest, celebrate this win, and prepare for tomorrow's brand new problem.

Build and deploy a model to predict employee attrition using features like job involvement, education, job satisfaction, and work life balance; treat it as classification evaluated with AUC and ROC.

Build, train, test, and deploy a no-code ml model in DataRobot to predict employee attrition from education, job involvement, satisfaction, and work life balance; Kaggle data fuels project card demo.

Explore the business case and success stories of ai in human resources, and seven ways ai transforms hiring, engagement, and analytics, with a reading and quiz on employee attrition.

Explore the data overview for the ai ml model, listing inputs like age, education level, job satisfaction, work life balance, and performance rating to predict attrition as a 0/1 classification.

Learn to upload employee data into DataRobot, perform initial exploration and feature visualization, select the target column, and begin training a binary attrition model using a drag-and-drop workflow.

Explore data in DataRobot by selecting employee attrition as the target, examine feature distributions, and set a 60/20/20 train-validation-holdout split for a classification model.

Train a machine learning model in DataRobot using a 60/20/20 split and review leaderboard results. Explore feature associations, like department with job role and monthly income.

Explore key classification model KPIs, including confusion matrix, precision, recall, and ROC curves, to evaluate accuracy, misclassification, and area under the curve on unbalanced data.

Assess model performance using AUC, confusion matrix, precision, and recall; explore feature importance and prediction explanations, then tune thresholds on ROC curves.

Deploy the best elasticnet classifier in DataRobot, build a predictor application, and test with multiple values, using average inputs to generate predictions and interpret explanations for employee attrition.

Celebrate completing day six of the no code ai bootcamp and prepare for tomorrow's brand new challenge with a good rest.

Train and deploy an image classifier to detect diabetic retinopathy with DataRobot, visualize activation maps of a convolutional neural network using grad-cam, and evaluate accuracy and recall.

Apply artificial intelligence, machine learning, and deep learning to healthcare by building, training, and deploying an image classifier to detect diabetic retinopathy with DataRobot.

Explore the business case for ai-driven diabetic retinopathy screening and success stories, as deep learning detects and classifies retinal images with high accuracy using DataRobot.

Explore image data to build an AI model for diabetic retinopathy classification using DataRobot. Analyze class distribution across five categories—healthy (no diabetic retinopathy) and four severity levels—highlighting an unbalanced dataset.

Upload your dataset to DataRobot with drag-and-drop, perform basic exploratory data analysis, and review class distributions to prepare training and testing data for model development.

Train a machine learning model in data robot by selecting target, assessing data quality, and evaluating models on leaderboard, with regularized logistic regression recommended for deployment and accuracy around 0.70.

Deploy the best model recommended for deployment by DataRobot using auto model, generate predictions from uploaded test data, and download a CSV with class probabilities for each image.

Explore explainable AI with gradcam visualizations to see how convolutional neural networks focus on image regions, using activation maps and heat maps to build trust and inform deployment.

Train an AI model to detect and classify chest x-ray images into healthy, COVID-19, bacterial pneumonia, and viral pneumonia, with Grad-CAM activation maps and DataRobot deployment.

Celebrate completing day seven and your milestone as you gain newly acquired skills. Rest, prepare for tomorrow's new challenge, and share feedback to improve future course launches.

Apply NLP to predict sentiment from Amazon Echo reviews using AI and ML. Build, train, and deploy a sentiment model; visualize text with word clouds and explore brand monitoring.

Train and deploy a sentiment analysis model from text data to predict customer happiness from Amazon Echo reviews, using data robot to train, test, deploy, and visualize common keywords.

Analyze the business case for an AI-powered sentiment analysis tool that predicts sentiment from text data, including Amazon Echo reviews, with NLP applications, reading materials, and a quiz.

Explore DataRobot to upload text data, perform exploratory data analysis and data quality checks, and train a sentiment analysis model that classifies reviews as positive or negative.

Analyze customer feedback data with DataRobot, exploring data quality, target selection on feedback, rating distributions, and word cloud insights while preparing models for training and holdout evaluation.

train and evaluate models in DataRobot, compare light gradient boosted trees to other models, and interpret AUC, ROC curves, confusion matrices, and feature importance for text data.

Deploy a no-code ai model with DataRobot, create an application and predictor, and run sentiment predictions through the deployed endpoint. Test positive and negative reviews to validate model behavior.

Drop the rating column to prevent data leakage and retrain on the no leakage Amazon reviews dataset; tune hyperparameters, analyze the confusion matrix, accuracy, and loss graphs, then deploy.

Complete final project part b by performing EDA and a word cloud on fake news dataset, assess performance, train with DataRobot, deploy the best model, and infer fake vs real.

Celebrate completing day eight, take a rest, and prepare for tomorrow’s brand new tool and challenge as we stay tuned for more.

Use AWS Sagemaker Autopilot to build, train, test, and deploy an AI model that predicts credit card default for bank customers, with hyperparameter tuning.

Use AWS SageMaker Autopilot to build, train, test, and deploy an AI model that predicts credit card default using customer features and outputs 0 or 1.

Explore five ai business applications—clustering, recommendations, classification, regression, and time series forecasting—using a Taiwan bank case with SageMaker Autopilot to predict credit card default and optimize marketing.

discover how AWS offers scalable cloud services for storage, compute, and AI tools. use S3, EC2, and SageMaker autopilot to train and deploy models without writing code.

Explore AWS S3 storage for uploading training data, EC2 for scalable compute, and SageMaker Autopilot for automated training, tuning, and deployment of ML models.

Explore AWS SageMaker Autopilot to upload credit card data via a Jupyter notebook, run a no-code script, train multiple models, compare metrics, and deploy the best model with explainability.

Configure an AWS SageMaker autopilot experiment with your S3 data and output locations and select the binary classification target. Explore preprocessing, feature engineering, hyperparameter tuning, and explainability results.

Explore AWS SageMaker Autopilot results from 250 trained models, compare accuracy and F1, review data exploration reports and feature importance, and deploy the best model for real-time predictions.

Delete the endpoint in AWS SageMaker Autopilot to avoid extra costs by navigating to inference endpoints, selecting the deployed endpoint, and choosing delete endpoint.

Predict health insurance charges as a regression task using age, gender, BMI, children, smoking, and region with AWS SageMaker Autopilot; train multiple models, select the best, analyze performance, and deploy.

Learn to set up an AWS SageMaker Autopilot experiment from your dataset in S3, upload data with a no-code script, run multiple model trials, and deploy the best model.

Explore no coding ai with AWS autopilot, preparing insurance_data.csv, setting the S3 data and outputs paths, selecting charges as target, and running auto-tuned regression models in a single experiment.

Learn to automate ai ml workflows with AWS Sagemaker autopilot, running experiments, tuning models for mse, deploying endpoints, and exploring data with candidate generation and data exploration notebooks.

Celebrate completing day nine of the bootcamp and stay tuned for a brand new tool tomorrow. Rest up and see you tomorrow.