Deep Learning & Machine Learning Masterclass w/ TensorFlowJS

Join this online course on Jaspin to learn HTML—the building blocks of the web—in a concise one-hour format with mobile-friendly viewing and short tutorials.

Discover how Jaspin enables real-time editing across devices, and learn the roles of HTML, CSS, and JavaScript—HTML for structure, CSS for styling, and JavaScript for functionality.

Learn how to set up a basic html document by declaring a doctype, using html and body tags, and understanding opening versus closing tags, with code coloring aiding productivity.

Explore proper body structure using header tags and paragraphs, from h1 to h6, with correct indentation and closing tags, and understand how SEO influences search results.

Apply text styles using style attributes to change color and background, combine multiple styles with semicolons, and experiment with fonts to create a negative space design.

Apply bold, italic, and underline tags in HTML, and manage paragraph tags with proper opening and closing to ensure clear, non-confusing styling.

Learn how to add a link in HTML using the a tag with an href attribute, insert the link text, and test navigation by clicking the hyperlink.

Learn to add an image to a web page by copying the image address, using a self-closing img tag, and adjusting width or height for proper display.

Add a link to an image by using an a tag with your image, making the image clickable to a website; remove extra breaks to keep the image properly formatted.

Learn to create unordered, ordered, and definition lists with list items and bullets, apply bold titles, adjust styling, and speed up coding with hotkeys ctrl+c and ctrl+v.

Explore how to create and structure tables, including borders, rows, cells, and headers. Learn why tables are used less today in favor of responsive web design.

Explore the different HTML input types, including text, password, color, email, number, date, radio, and checkbox, and learn how names and strings reference inputs.

Html forms the foundation, while JavaScript adds functionality; styles and scripts now live in separate css and JavaScript files, organized in the same project folder.

Apply inline styles to HTML elements using the style attribute, covering color, background-color, font-family, and font-size. Understand opening and closing tags, self-closing tags, and moving styles to a separate file.

Apply CSS styling to HTML by defining h1 and p elements with color, background color, font-family, and font size using curly brackets and semicolons, while understanding defaults and error cues.

Use unique IDs to differentiate multiple paragraphs and apply CSS selectors to override base styles. See how #first paragraph and #second paragraph change color, font, and background with case sensitivity.

Apply CSS classes to multiple tags to style paragraphs consistently, reuse classes across elements, and implement global changes such as font color or italics, with changes propagating site-wide.

Explore font families by comparing serif fonts, sans serif, cursives, and model space fonts, and learn how readability and consistent widths influence web and print design.

Learn to use RGB values (red, green, blue) from 0-255 with alpha transparency, create off-black and off-white palettes, and apply hex codes for CSS color design.

Learn how to apply color in CSS using RGB values and hex codes, explore off-black and off-white tones, and understand RGB, alpha, hue, saturation, and brightness for crisp UI design.

Explore text align options (left, center, right, justify) and text transforms (uppercase, lowercase, capitalize) to style headers and pages, emphasizing page aesthetics and bold, uppercase headings.

Master responsive design by setting margins, widths, and padding to keep content centered across devices, and balance percentage and fixed widths with relative and absolute positioning.

Learn how to set the body width and margins for a webpage, adjust it to 60–70 percent, apply background and text color changes, and use negative space to improve readability.

Explore how a Latin text generator provides placeholder content to build a simple web page, showing how titles, body text, paragraphs, and padding enhance layout.

Create a horizontal menu bar with an unordered list and active home link, then style margins, padding, overflow hidden, hover colors, and uppercase Verdana text for navigation.

Add a background image to your page by selecting a public domain image, copying its address, and applying it to the body with no-repeat and background-attachment: fixed.

Experiment with margins and padding in css to center content, test large and negative values, and use complementary colors (orange vs blue) to create visual contrast.

The one-hour course builds a foundation in deep learning and machine learning with TensorFlowJS, starting from zero JavaScript knowledge through efficient, short tutorials and semi scripted live coding with Jasmine.

Learn how HTML builds structure, CSS styles, and JavaScript adds interactivity, using a paragraph tag and color styling. See the console and alert usage, with semicolon-terminated JavaScript lines printing strings.

Explore how to scale functions by creating multiple focused functions, naming them clearly (print name, update score, calculate), and organizing code for web apps and games.

Declare and initialize variables to hold numbers and strings, like variable number and variable string. See how JavaScript blends data types, concatenates strings with numbers, and updates values with addition.

Learn how arrays work in JavaScript, including indexing, square bracket syntax, and handling undefined when out of range, with examples of number and string arrays.

Understand variable scope by contrasting global variables that span the page with variables inside curly brackets that are limited to that block, as explained for JavaScript.

Learn how to add a text input in HTML, capture user input with a variable, and update the display using document.getElementById, illustrating basic input handling in JavaScript.

Call JavaScript functions from button clicks by assigning an id and using an on click handler to change text; call functions from other functions to avoid repetition and improve efficiency.

Learn how if statements use a condition to trigger actions in code, with comparisons like equal to, not equal to, less than or equal to, and greater than.

Master if, else if, and else blocks to control program flow and handle specific conditions. Follow best practices by reserving else for fallback cases and avoiding overlap.

Learn to change styles via JavaScript by wiring buttons to update color, font size, and font family using document.getElementById and element.style, and updating innerHTML on click.

Discover what machine learning is and how data drives improvements in software. Explore examples like data mining, search customization, and autonomous systems, plus the supervised, unsupervised, and reinforcement branches.

TensorFlow.js brings machine learning to JavaScript, in browser or Node.js. Load via CDN or Node.js, access features like cameras and microphones, and import Python models for flexible web projects.

Load TensorFlow.js into your web project via a CDN or Node.js, verify the tf object in the console, and prepare to build your first machine learning model with TensorFlow.

Explore core linear algebra concepts for machine learning, including scalars, vectors, matrices, and tensors, and learn how tensors enable GPU-accelerated data processing with TensorFlow.

Build and explore tensors with TensorFlow.js by creating 1d, 2d, and multi-dimensional tensors, and inspect rank, shape, and dtype while viewing contents with print.

Explore tensor utility methods in the Tenzer Flow JS library, including reshape, ones, zeros, and truncated normal, and see how reshaping and prints reveal tensor dimensions for machine learning data.

Explore core tensor math by adding blue and red tensors, computing dot products, and transposing matrices; understand axes as rows and columns and prepare for visualization with TensorFlow.js.

Explore deep learning, a multilayered subset of machine learning and artificial intelligence, driven by data, powerful processors, and innovative algorithms, with larger neural networks yielding higher performance than traditional models.

Learn one hot encoding and its use in neural networks, with a hands-on project that classifies four classes—mammal, fish, bird, and reptile—using one hot labels.

Set up training data using one-hot encoding to map eight animal classes to eight-by-eight input tensors and four-class labels, preparing tensors train X and train Y for model building.

Build a neural network with TensorFlow.js by stacking two dense layers (30 relu, 4 sigmoid) for eight inputs and four classes, then compile with an optimizer and loss.

Train your neural network with fit on training data (train X, train Y) using batch size 64 and multiple epochs (30–200), tracking loss and accuracy via epoch logs.

Build training data for detecting lines in 4x4 grayscale images, with eight one-hot labeled samples and tensors shaped 8x4x4 for X and 8x8 for Y in a Tenzer Flow workflow.

Build a convolutional neural network in TensorFlow's sequential model with conv2d layers on 4x4x1 input, flatten to dense eight outputs for one-hot labels, trained with Adam and mean squared error.

Train a convolutional neural network through a for loop of epochs, fit on training data (train X and train Y), and monitor epoch loss to understand learning and overfitting.

Predict the number of lines in a four by four grayscale image by feeding a one-hot encoded input to a convolutional neural network and testing with tensor data.

Explore what a recurrent neural network is and how long short term memory enables context-aware language tasks like speech recognition, translation, and reading comprehension.

Generate a sequence and label for an LSTM, detecting repeating subsequences of at least three to assign labels one or zero and build a training dataset.

Generate a dataset by creating random sequences of a given length and sample count, store sequences and binary labels in TensorFlow buffers, and return the prepared data for model training.

Construct a TensorFlow sequential neural network with an eight-unit LSTM layer and a dense output, using binary cross entropy loss and Adam for training.

Train the neural network by generating a dataset and fitting with X_train and Y_train across epochs, then monitor accuracy and validation accuracy while tuning learning rate and sigmoid activation.