Lecture

Interpretable Machine Learning: Sparse Decision Trees and Interpretable Neural Networks

Graph Chat

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

Description

This lecture by the instructor covers the extremes of interpretability in machine learning, focusing on sparse decision trees and interpretable neural networks. The presentation delves into the technical definition of interpretable machine learning models, the problem spectrum of tabular and raw data, the comparison between COMPAS and CORELS models, and the prediction accuracy of re-arrest within 2 years. It also explores the optimal sparse decision trees construction, dynamic programming for branch and bound, and the Generalized and Scalable Optimal Sparse Decision Trees (GOSDT). Additionally, the lecture discusses interpretable neural networks, the challenges of explaining deep neural networks, and the development of an interpretable AI algorithm for breast lesions. The session concludes with a summary of modern decision tree methods and the existence of interpretable neural networks for computer vision.

Official source

Related concepts (32)

Decision tree learning

Decision tree learning is a supervised learning approach used in statistics, data mining and machine learning. In this formalism, a classification or regression decision tree is used as a predictive model to draw conclusions about a set of observations. Tree models where the target variable can take a discrete set of values are called classification trees; in these tree structures, leaves represent class labels and branches represent conjunctions of features that lead to those class labels.

Decision tree

A decision tree is a decision support hierarchical model that uses a tree-like model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility. It is one way to display an algorithm that only contains conditional control statements. Decision trees are commonly used in operations research, specifically in decision analysis, to help identify a strategy most likely to reach a goal, but are also a popular tool in machine learning.

Recurrent neural network

A recurrent neural network (RNN) is one of the two broad types of artificial neural network, characterized by direction of the flow of information between its layers. In contrast to uni-directional feedforward neural network, it is a bi-directional artificial neural network, meaning that it allows the output from some nodes to affect subsequent input to the same nodes. Their ability to use internal state (memory) to process arbitrary sequences of inputs makes them applicable to tasks such as unsegmented, connected handwriting recognition or speech recognition.

Artificial neural network

Artificial neural networks (ANNs, also shortened to neural networks (NNs) or neural nets) are a branch of machine learning models that are built using principles of neuronal organization discovered by connectionism in the biological neural networks constituting animal brains. An ANN is based on a collection of connected units or nodes called artificial neurons, which loosely model the neurons in a biological brain. Each connection, like the synapses in a biological brain, can transmit a signal to other neurons.

In chemistry, pH (piːˈeɪtʃ ), also referred to as acidity, historically denotes "potential of hydrogen" (or "power of hydrogen"). It is a scale used to specify the acidity or basicity of an aqueous solution. Acidic solutions (solutions with higher concentrations of hydrogen () ions) are measured to have lower pH values than basic or alkaline solutions. The pH scale is logarithmic and inversely indicates the activity of hydrogen ions in the solution. where [H+] is the equilibrium molar concentration (mol/L) of H+ in the solution.

Related lectures (64)

Nonlinear Supervised Learning

Explores the inductive bias of different nonlinear supervised learning methods and the challenges of hyper-parameter tuning.

Ground Penetrating Radar: Data Analysis

Explores the automated picking of reinforcement bars within Ground Penetrating Radar data using machine learning and signal processing techniques.

Reinforcement Learning Concepts

Covers key concepts in reinforcement learning, neural networks, clustering, and unsupervised learning, emphasizing their applications and challenges.

Perception: Data-Driven Approaches

Explores perception in deep learning for autonomous vehicles, covering image classification, optimization methods, and the role of representation in machine learning.

Machine Learning for Feature Extraction

Explores machine learning for feature extraction, 3D vision, and neural networks in mobile robotics.

Related concepts (32)

Decision tree learning

Decision tree

Recurrent neural network

Artificial neural network