Colin Batchelor retweeted
#AI4SD2022 Dr Jo Grundy of @ECSUoS's talk on: Outlier Detection in Scientific Data is now online: youtu.be/MhaJ2IZ2IBc #ArtificialIntelligence #AI #ML #MachineLearning #Data #Outliers #DataScience @dotmatics @BSI_UK @Patterns_CP @CASChemistry @digital_rsc @ErrantScience
After at least five years of using RStudio I have only just noticed you can edit .gitignore directly in it without using vi in the Terminal pane.
Name reactions: not just for first-year undergraduate organic chemistry.
🎉✨Luca Banfi et al. and The 100 facets of the Passerini reaction feature this week on our first #ChemSciCovers, at the front!
🔗ow.ly/6vgG50H6PBh
Last call for job going in my team: careers-rsc.icims.com/jobs/2…
I will have something to broadcast shortly but apparently your stream of tweets should be 4 "engagements" to 1 "broadcast".
Colin Batchelor retweeted
Verge article describes incoming STEM students not being familiar with file systems — a problem I encountered while teaching at #royalholloway that led me to make this: superbasics.beholder.uk/file…
Colin Batchelor retweeted
Our second talk in #AI3SDSemanticWebTechnologies is now online: youtu.be/Bl-8V48hjgc! @documentvector from @RoySocChem speaking on: Ontologies, Natural Language, Annotation & Chemistry! #Ontologies #SemanticWeb #LinkedData #Semantics #RDF #OWL #NLP #Annotation
I am speaking at this event next Wednesday: ai3sd-winter-series-100221.e…
Colin Batchelor retweeted
We may be locked down but AI3SD has many events to keep you entertained! Register 4 #AI3SDML4ScientificDiscovery 20/01: ai3sd-winter-series-200121.e… Speakers: Professor Jacqui Cole @MolecularEngine & Dr Jo Grundy @ECSUoS #ML #MachineLearning #ArtificialIntelligence #AI #DataScience
Colin Batchelor retweeted
Register 4 #AI3SDSemanticWebTechnologies: 10/02: #SemanticWeb Technologies in #Chemistry ai3sd-winter-series-100221.e… Speakers: @SamiKanza/@unisouthampton, @documentvector/@RoySocChem & Alexandra Simperler of Goldbeck Consulting (@SimpConsulting @gerhardgoldbeck) #RDF #OWL #Ontologies
Come and work in my team: careers-rsc.icims.com/jobs/2…
Colin Batchelor retweeted
Register for our 4th #AI3SD Winter Seminar: 20/01: ML 4 Scientific Discovery ai3sd-winter-series-200121.e… Speakers: Professor Jacqui Cole of @MolecularEngine and Dr Jo Grundy of @ECSUoS! #ML #MachineLearning #ArtificialIntelligence #AI #ScientificDiscovery #Data #DataScience
YOUR TREEMAP BOXES ARE ALL THE WRONG SIZE STOP IT
Replying to @RishiSunak
12/ The funding for @beisgovuk will cement the UK as a science superpower through R&D and innovation spending, continue to support businesses through access finance schemes and stimulate the economic recovery post Covid. #SR20
Colin Batchelor retweeted
My new book, Artificial Intelligence in Drug Discovery, is published today. You can read a preview here, but using TXTBK25 on the @RoySocChemBooks website gives you a 25% discount. smile.amazon.co.uk/dp/B08MYS… #CompChem #RealTimeChem #ArtificialIntelligence #MachineLearning #AI #ML
Colin Batchelor retweeted
Out now! Artificial Intelligence in Drug Discovery by @nathanbroon
A wealth of information from leading experts in the field and ideal for students, postgraduates and researchers in industry and academia.
👀📚: rsc.li/2JP0QjD
#artificialintelligence #RSCBooks
On the one hand, it's good to have a collection of NLP datasets in the one place, on the other hand I've never seen the Bender rule broken so often on a single web page: huggingface.co/datasets (@huggingface - please sort this out).
Colin Batchelor retweeted
Month-on-month change charts for economic indicators should be banned in a context like this.
A 20.3% contraction followed by a 1.8% expansion is *anything but* a V-shaped recovery, yet the at-a-glance takeaway here is V.
There's just no need to make this chart.
Colin Batchelor retweeted
The steepness of the line should be an immediate red flag: there’s no way something like that can be genuine, so it must come from a quirk in the data.
What quirk? This quirk xitter.cf/jburnmurdoch/status/12…
Replying to @jburnmurdoch
Happily, rather than just subtracting 30k tonight and leaving the old data unchanged, the full time series has been revised, so we now appear to have consistent data on cases by specimen date right back to February (for the UK, England, and English subnational areas).


















