Deep Learning for Non-Programmers

Description:

This course provides a fun and non-technical introduction to Deep Learning. Students will get conceptual understanding of how unstructured data such as text, images data are understood by machines to perform tasks such as image and text classification and clustering. There will be a section on understanding how to handle time series data. Students will build different types of deep learning models (without writing a single line of code). Students will be using the Orange Data Mining tool to perform hands-on activities.

Duration: 1 Day

Course Code: BDT288

Learning Objectives:

After this course, you will:

• Gain an introduction to deep learning, understanding its fundamentals, tools, and techniques.
• Learn how to work with images, performing clustering and classification to predict image content.
• Explore working with text, creating embeddings, and conducting sentiment analysis and clustering.
• Develop skills in handling time series data, building models for future value prediction.
• Master the art of visualizing geo data using latitude and longitude tools.

• Must have taken course: Artificial Intelligence and Machine Learning Basics for Non-Programmers

• This course is designed for anyone interested in working on unstructured data such as text and images to perform classification, clustering, visualizations.

• This course is designed for anyone interested in working on unstructured data such as text and images to perform classification, clustering, visualizations.

Course Outline:

1. Introduction to Deep Learning

• Understanding the basics of artificial intelligence and machine learning
• Predictive and Generative Models
• Overview of neural networks and deep learning concepts
• Applications of deep learning in text, image, and time series data
• Demo: Using neural networks

2. Working with images

• Introduction to image data and image representation
• Leveraging knowledge learnt by community to handle images (transfer learning)
• Using Orange’s deep learning toolbox for image classification
• Hands-on: Building multiple models to perform binary and multi-class classification of images

3. Working with Text

• Introduction to text data and its different uses in deep learning
• Preprocessing text data for deep learning tasks
• Understanding word embeddings and text representations in Orange
• Hands-on: Learn how to build model to perform sentiment analysis on movie review data using couple of word embedding techniques

4. Handling Time Series data

• Understanding time series data and its unique characteristics
• Leveraging different prebuild models in Orange to perform time series forecasting
• Hands-on: Build model to forecast Google stock price based on historical data

5. Visualizing Geo Data

• Understanding what Geo location data
• Using Orange to quickly visualize data based on Geo location
• Hands-on: Use Orange to visualize data based on Latitude and Longitude

Training material provided: Yes (Digital format)

Hands-on Lab: Students must install Orange Data Mining software.