Interpretable machine learning applications: Part 3

via Coursera

Overview

Title: Interpretable Machine Learning Applications: Part 3

Description: Join our intensive 50-minute project-based course, titled "Interpretable Machine Learning Applications: Part 3", available on Coursera. This course is designed to equip participants with the ability to apply sophisticated explanation techniques and algorithms for interpreting predictions made by complex machine learning models, such as artificial neural networks. The technique utilized is ideal for analyzing models that function as 'black-boxes', making this course incredibly beneficial for decision-makers in various sectors, including banking and public administration. These explanations are pivotal for those wishing to employ trusted machine learning applications to enhance decision-making processes.

The course is structured around three core learning objectives:

Objective 1: Define, train, and evaluate an artificial neural network (Sequential model) based classifier using the Keras API for TensorFlow. Participants will train and test the predictive model using the HELOC dataset, which includes data on approved and rejected mortgage applications.
Objective 2: Learn to generate explanations by examining profiles similar to a mortgage applicant categorized as either a "Good" or "Bad" risk.
Objective 3: Generate contrastive explanations focusing on feature and pertinent negative values to determine what changes are needed to convert a "rejected" application into an "approved" one.

Provider: Coursera

Categories: Machine Learning Courses