Mediaspace scheduled maintenance: Aug 25, 2026 07:00 - 12:00 AM. During this time, videos will be temporarily unavailable. Check status updates.

Lecture

Data Driven Science: MODNet Methodology

Description

This lecture covers the MODNet methodology, a supervised machine learning framework for material property predictions based on composition or structure. It includes feature selection, genome encoding, property grouping, and property prediction. Results show significant gains in accuracy and the importance of frameworks for small datasets.

Official source

Related concepts (32)

Machine learning

Machine learning (ML) is an umbrella term for solving problems for which development of algorithms by human programmers would be cost-prohibitive, and instead the problems are solved by helping machines 'discover' their 'own' algorithms, without needing to be explicitly told what to do by any human-developed algorithms. Recently, generative artificial neural networks have been able to surpass results of many previous approaches.

Materials science

Materials science is an interdisciplinary field of researching and discovering materials. Materials engineering is an engineering field of finding uses for materials in other fields and industries. The intellectual origins of materials science stem from the Age of Enlightenment, when researchers began to use analytical thinking from chemistry, physics, and engineering to understand ancient, phenomenological observations in metallurgy and mineralogy. Materials science still incorporates elements of physics, chemistry, and engineering.

Feature selection

Feature selection is the process of selecting a subset of relevant features (variables, predictors) for use in model construction. Stylometry and DNA microarray analysis are two cases where feature selection is used. It should be distinguished from feature extraction. Feature selection techniques are used for several reasons: simplification of models to make them easier to interpret by researchers/users, shorter training times, to avoid the curse of dimensionality, improve data's compatibility with a learning model class, encode inherent symmetries present in the input space.

Feature engineering

Feature engineering or feature extraction or feature discovery is the process of extracting features (characteristics, properties, attributes) from raw data. Due to deep learning networks, such as convolutional neural networks, that are able to learn it by itself, domain-specific- based feature engineering has become obsolete for vision and speech processing.

Feature learning

In machine learning, feature learning or representation learning is a set of techniques that allows a system to automatically discover the representations needed for feature detection or classification from raw data. This replaces manual feature engineering and allows a machine to both learn the features and use them to perform a specific task. Feature learning is motivated by the fact that machine learning tasks such as classification often require input that is mathematically and computationally convenient to process.

Related lectures (52)

Machine learning: Physics and Data

Delves into the intersection of physics and data in machine learning models, covering topics like atomic cluster expansion force fields and unsupervised learning.

Applied Machine Learning

Introduces applied machine learning concepts such as data collection, feature engineering, model selection, and performance evaluation metrics.

Active Learning for Molecular Design

Covers machine learning approaches for material design, practical examples, and software tools for research.

Machine Learning Biases

Covers the basics of machine learning, challenges in deployment, adversarial attacks, and privacy concerns.

Principal Component Analysis: Dimensionality Reduction