learn/what the ai was for
Property prediction
Property prediction is the use of a trained computer model to guess a measurable quantity about something that already exists, rather than measuring it in the laboratory or telescope.
What it is
A property is anything about an object that could in principle be measured: how hot a chemical reaction needs to be, how hot a star's surface is, how stable a protein is, whether a bump in a signal is real. Property prediction takes a description of the object that is already cheap to obtain, and maps it to the quantity of interest. The map is learned from past cases where both the description and the measured answer are known. The model is shown many such pairs, adjusts itself until its guesses match the recorded measurements, and is then applied to objects whose answer has not been measured.
The description fed in can be almost anything a computer can read: a list of the elements in a material, a protein's sequence of building blocks, or the raw light spectrum of a star. The output is usually a single number, a set of numbers, or a yes-or-no label. Nothing about the object is changed or invented; the object is fixed and only the measurement is being guessed. This separates property prediction from design, where the task is to invent a new object with a desired property. Here, the object comes first and the number follows.
Why scientists use it
Measurement is often the slowest and most expensive step in a science. Growing a crystal, running a chemical synthesis, recording a high-quality spectrum of a faint star, or assaying a protein in a tube all cost time, money and instrument access. If a model can supply a usable estimate in a fraction of a second, a researcher can sort thousands of candidates and measure only the promising few, or can attach numbers to observations that already sit in an archive and would otherwise go unused.
Property prediction therefore stands in for an instrument. In some projects it fills a gap in a pipeline, supplying a setting or a filter so that work can carry on without pausing for a measurement. In others it transfers careful measurements made on a small, well-studied sample to a much larger sample that was observed more cheaply. The model is not a new theory of the system; it is a shortcut through a measurement that someone, somewhere, has already made many times.
In the record
Limits
A learned map only reflects the cases it was trained on. Shown an object unlike anything in that set, it still returns a confident-looking number, with no warning that it is guessing. Errors in the original measurements are copied into the predictions. The model also finds patterns that happen to hold in its training data, not causes, so it cannot explain why a property takes the value it does, and it will mislead if the data were gathered in a skewed way. Predictions narrow what to measure; they do not replace measurement.
Written by AIxSci: the background from general knowledge, the examples from checked records.
By field
- Property prediction in structural biology and protein science
- Property prediction in materials and computational chemistry
- Property prediction in astronomy and astrophysics
- Property prediction in weather, climate and oceans
Every article
aix-00286
AI fills gaps in river temperature records to track overlapping heatwaves
aix-00288
Machine learning turns satellite cloud images into hourly rainfall maps for central Vietnam
aix-00287
Satellite image dataset extended to four channels for typhoon strength prediction
aix-00289
Machine learning points to filter loading as main driver of levoglucosan desorption temperature
aix-00281
Bacteria-derived aerosol matter made to produce oxidants, then modelled from its optical fingerprints
aix-00290
Machine learning separates weather from policy in China's winter heating air pollution
aix-00293
Machine learning used to predict fine particle emissions at a Greek regional airport
aix-00280
Neural network splits forest carbon measurements into uptake and release
aix-00279
Machine learning predicts where dopant atoms sit in hexaferrite crystals before refinement
aix-00275
AI model trained on protein folding predicts how mutations affect binding
aix-00272
Tree-based models predict sliding wear of carburised and plain engineering steels
aix-00271
Machine learning predicts galaxy 'relicness' from survey measurements, cutting spectroscopy demand
aix-00218
Sparse regression fits crystal force constants to model heat flow and vibrations
aix-00219
Neural network supplies the sideways flows needed to measure the quiet Sun's magnetic energy
aix-00220
Neural network predicts how titanium alloy grains round off during annealing
aix-00221
Neural network measures rotation periods for Kepler's main-sequence stars
aix-00222
Neural networks trained on molecule pairs predict light absorption in stacks of fifty
aix-00223
Neural networks weigh cluster X-ray maps to estimate galaxy cluster masses
aix-00216
Neural network models simulate how lead titanate loses its electrical polarisation on heating
aix-00213
Neural networks predict how much ultrasound energy drives magnesia dissolution
aix-00254
Neural networks read a fast radio burst's dispersion straight from its spectrum
aix-00210
Language models suggest ingredients and firing temperatures for inorganic materials recipes
aix-00209
Neural network turns Kaguya camera images into global maps of lunar surface chemistry
aix-00208
Neural network predicts atomic charges from bond lists, without atomic coordinates
aix-00206
Neural network designs five hard metallic glasses, all confirmed by experiment
aix-00205
Machine learning fitted to flexible molecules predicts oxidation potentials and hydration energies
aix-00159
Machine learning assigns atomic charges to size up polarisation in bismuth vanadate conductors
aix-00158
Structure predictions place an uncharacterised human protein in the BRICHOS family
aix-00157
Microscopy and a trained classifier map which E. coli proteins clump together
aix-00155
Neural network predicts crystal density of candidate explosives from molecular structure
aix-00154
AlphaFold models used to map membrane-crossing segments in human proteins
aix-00150
Deep learning sorts crystal structures for single-molecule magnet behaviour
aix-00147
Testing whether symmetry-aware neural networks can tell apart atomic arrangements in perovskites
aix-00160
Neural networks predict how zeolites take up carbon dioxide from structure alone
aix-00163
Peptides designed by simulation and machine learning sit at condensate surfaces
aix-00169
AlphaFold's confidence scores used to predict which disordered protein segments bind LC8
aix-00170
Machine learning predicts strength of PLA plastic filled with boron nitride flakes
aix-00172
Machine learning tested against three definitions of protein shape-shifting
aix-00146
Machine learning ranks stable materials by predicted superconducting temperature
aix-00139
Robot laser-heats thin films while software picks each next heating condition
aix-00173
Machine learning predicts nanofibre thickness from electrospinning settings, tested against new scaffolds
aix-00178
Machine learning picked catalyst recipes for a loop of 44 lab cycles
aix-00133
Regression models predict how graphene loading changes aluminium's electron emission
aix-00132
Five codes measured galaxy shapes in simulated Euclid images, one using neural networks
aix-00131
Machine-learned potential predicts how chromium sulfide layers rearrange during exfoliation
aix-00179
Neural network predicts protein pair binding strength from simplified molecular models
aix-00180
Teaching protein language models to rank mutants from a few dozen lab measurements
aix-00183
Neural network predicts nanostructure optics to design a two-colour laser collimator
aix-00186
Language model trained on protein sequences predicts shape and stability of disordered proteins
aix-00187
Machine-learned force field used to model two crystal forms of formamide
aix-00191
Neural network stands in for slow spectral modelling to read 64 stars' chemistry
aix-00129
Neural networks sort supernovae and estimate their distances from brightness alone
aix-00124
Software predictors compare protein floppiness across retinal disease protein sets
aix-00123
Neural networks swap places with black hole image simulations, both directions
aix-00117
Deep learning model predicts polymer heat-softening across 48,208 designed candidates
aix-00114
A compact way to describe molecules speeds up machine learning of their properties
aix-00111
Zero-padding lets one neural network encode crystals with differing element counts
aix-00110
Deep learning trained on robotic peptide mapping predicts where antibody proteins degrade
aix-00108
Oxidation potentials for 15,238 molecules assembled from quantum calculations and lab measurements
aix-00103
Neural network predicts yield of chitosan nanoparticles grown with olive leaf extract
aix-00101
Simulations map three sodium–bismuth compounds, then models predict their heat-to-electricity performance
aix-00099
Weighing the Milky Way's disk reveals a mass excess along the Local Arm
aix-00097
Neural networks speed up forecasts of what gravitational wave detections LISA will catch
aix-00095
Neural networks estimate dark matter halo masses from a galaxy's neighbours
aix-00090
Machine learning reads planet-forming disk masses from archival ALMA observations
aix-00089
Neural networks trained on atomic surroundings score how mutations change proteins
aix-00088
Neural network retrained to predict how crystals absorb light, from few examples
aix-00084
Neural networks read simulated galaxy cluster catalogues to estimate cosmological parameters
aix-00080
Machine learning sifts three-colour infrared survey data for cold brown dwarfs
aix-00077
Neural networks sift 812,118 quasar spectra for hidden gravitational lenses
aix-00074
Software predictions map floppy stretches of touch-sensing PIEZO channels
aix-00071
Simulations map how surface atoms rearrange in alloys when molecules attach
aix-00070
Neural network scores candidate composite microstructures in a strength–toughness design loop
aix-00069
Neural networks rebuild cosmic expansion history to calibrate gamma-ray burst energies
aix-00068
Machine learning predicts marine steel corrosion from seawater conditions and alloy make-up
aix-00065
Model predicts how fast an enzyme works on a substrate, with uncertainty attached
aix-00063
Adding a term for the unfolded protein sharpens AI stability predictions
aix-00058
Deep-learning model designs petrol blends from octane and soot targets
aix-00056
Deep learning tool predicts how single mutations weaken protein partnerships
aix-00051
Ultrasound-treated clay bleaches cooking oil, with machine learning picking the settings
aix-00046
Two neural networks predict how heat shrinks the band gap of silver-based crystals
aix-00045
Neural network reads simulated quasar light curves to estimate black hole properties
aix-00044
Machine-learnt electron densities drive molecular dynamics of a gold–saltwater interface
aix-00040
Statistical model of 5,441 perovskite solar cells predicts efficiency and suggests recipes
aix-00035
GPT-4 reads band gap values from paper sentences to train better predictors
aix-00033
Neural network built from physics equations extracts solder deformation coefficients
aix-00024
Machine learning picked which engineered enzymes to build for fatty alcohol production
aix-00021
Neural network reads a million star spectra to measure temperature and chemistry
aix-00016
Machine learning picked which shape-memory alloys to make next, round by round
aix-00015
Transformer model predicts how gas-storing crystals take up gases, replacing slow simulations
aix-00011
A learning algorithm ran a synchrotron beamline to find a phase-change material
aix-00010
Robotic lab with machine learning proposes and runs 353 inorganic synthesis experiments
aix-00009
Gaussian process models pick light-sensitive ion channels that reach mammalian cell membranes
aix-00007
Neural network predicts protein atom positions from amino acid sequence alone
aix-00004
Robot lab and learning agent redesign an enzyme to survive higher temperatures
aix-00002
Deep learning turns raw NMR spectra into protein structures without human intervention
aix-00001